← Back

Picolinate ruthenium(II)-arene complex with in vitro antiproliferative and antimetastatic properties: comparison to a series of ruthenium(II)-arene complexes with similar structure.

PMID: 22266461
JournalofInorganicBiochemistry108(2012)53–61 ContentslistsavailableatSciVerseScienceDirect Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio Picolinate ruthenium(II)–arene complex with in vitro antiproliferative and antimetastatic properties: Comparison to a series of ruthenium(II)–arene complexes with similar structure Nevenka Gligorijevića, Sandra Aranđelovića, Lana Filipović a, Ksenija Jakovljević a, Radmila Janković a, Sanja Grgurić-Šipka b, Ivanka Ivanović b, Siniša Radulovića,⁎ , Živoslav Lj. Tešić b aInstituteforOncologyandRadiologyofSerbia,Pasterova14,11000Belgrade,Serbia bFacultyofChemistry,UniversityofBelgrade,Studentskitrg12-16,11000Belgrade,Serbia a r t i c l e i n f o a b s t r a c t Articlehistory: Inourpreviousstudy,ruthenium(II)-p-cymenecomplexesofgeneralformula[(η6-p-cymene)Ru(L)Cl2],L:3- Received18July2011 acetylpyridine (1), 2-amino-5-chloropyridine (2); and [(η6-p-cymene)Ru(HL)Cl], HL: 2,3-pyridinedicar- Receivedinrevisedform8December2011 boxylicacid(3),2,4-pyridinedicarboxylicacid(4),revealedlowantiproliferativeactivity,exceptcomplex Accepted9December2011 [(η6-p-cymene)RuCl(picolinicacid)]·H O(5)whichexhibitedIC around80μM.Inthisstudywefurtherin- Availableonline16December2011 2 50 vestigatedinvitropotentialofantimetastaticactionofrutheniumcomplexesonHeLaandtwoendothelial celllines.Comparisonofstructureandactivityoffivecomplexes indicatedheterogenicmode ofactivity, Keywords: Ruthenium(II)–arene withregardtothepotentialofantimetastaticandantiproliferativeeffect.Replacementofsubstitutedpyri- Antiproliferativeactivity dineligandwithpicolinicacid(complex5)aroundRu(II)centercontributedtocomplexcytotoxicityandru- Antimetastatic theniumDNAbindingaffinity.Analysisofruthenium(II)accumulationinDNAandproteinfractionsofHeLa ERCC1 cells,usingICP-OESrevealedsignificantlyhighercontentofcomplex5inDNAfractionincomparisontothe PARP othertestedcompounds.Italsoalteredcellcycleprogression,affectedexpressionofDNArepairenzymes ERCC1andMSH2,andshowedenhancedactivityincombinationwith3-aminobenzamide. Regardlessoftheireffectoncellgrowth,Ru(II)complexesexertedantimetastaticeffectonseveraltumorcell linesinvitro,achievedmostlybytheeffectoncelladhesion,migrationandangiogenesis,whilepicolinate ruthenium(II)–areneadditionallyexertedinhibitoryeffectonextracellularmatrixdegradation. ©2011ElsevierInc.Allrightsreserved. 1.Introduction activity [1]. Interestingly, some ruthenium complexes demonstrate greater efficacy against cancer metastasis than against primary tu- Cisplatinandstructurallyrelatedplatinumbaseddrugsrepresent mors. Examples of such complexes include the two complexes unique and important class of antitumor agents. Among the great NAMI-AandKP1019whichunderwentphaseIIclinicaltrials[1]. number of non-platinum metal complexes examined for anticancer Themostnumerousgroupofcytotoxicrutheniumcompoundsis characteristicsuptodate,recentattentionisfocusedonruthenium Ru(II)–arenecomplexes,whichweredevelopedprimarilybyDyson complexes.Rutheniumcomplexesarereportedtoshowpatterns of andco-workersandSadlerandco-workers,althoughnoneofthese antitumor activity and clinical toxicity that are distinct from those compoundshasyetenteredclinicaltrials[5]. ofplatinumcomplexes[1–4]. Inourpreviousstudiesseriesofruthenium(II)complexescontaining The most likely interactions of ruthenium complexes are with: thep-cymeneligandandapyridinederivativecoordinatedinamono- plasma proteins, collagens of the extracellular matrix (ECM),actins dentate or bidentate manner were synthesized and characterized by on the cell surface, regulatory enzymes within the cell membrane standardanalyticaltechniques[6,7]. and/orinthecytoplasmandDNAinthecellnucleus,whichcangive Analysis of cell growth inhibition caused by Ru complexes of them diversity of modes of action involving both extra- and intra- structural formula [(η6-p-cymene)Ru(L)Cl ], (L: 3-acetylpyridine 2 cellularprocesses[5]. (1), 2-amino-5-chloropyridine (2)); and [(η6-p-cymene)Ru(HL)Cl], In the last 30years, a large number of ruthenium containing (HL:2,3-pyridinedicarboxylicacid(3),2,4-pyridinedicarboxylicacid agents have been synthesized and tested for potential anticancer (4)), and [(η6-p-cymene)RuCl(picolinic acid)]·H O (5) (Fig. 1) on 2 panel of tumor cell lines (HeLa, B16, MDA-MB-361, MDA-MB-453 and LS-174) revealed low antiproliferative activity (IC >200μM), 50 ⁎ Correspondingauthor.Fax:+381112067294. exceptforcompound5withIC 50 81.97μMasalreadyreported[6,7]. E-mailaddress:sinisar@ncrc.ac.rs(S.Radulović). KnowingthatfewRucomplexeswhichfailedtraditionalanticancer 0162-0134/$–seefrontmatter©2011ElsevierInc.Allrightsreserved. doi:10.1016/j.jinorgbio.2011.12.002 54 N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 Fig.1.Structuresofinvestigatedruthenium-p-cymenecomplexes. drugscreeningprogramsbasedontheirlevelofgenotoxicityandcy- DNA-repair-dependent signaling which includes components of totoxicity in vitro, have demonstrated antimetastatic activity [1–3], NERorMMRisutilizedincellresponsetoruthenium(II)-p-cymene wefurtherinvestigatedmechanismofactionofourRu(II)-p-cymene complexes, by following expression of ERCC1 (mRNA and protein complexes, and extended our research on endothelial cells lines: level)andMSH2(proteinlevel)usingQuantitativeReal-TimePCR EA.hy926(transformedhumanumbilicalveinendothelialcells)and (RQ-PCR) and Western blot. PARP is another nuclear enzyme re- MS1(murineendothelialcells),asmodelsystemsforinvitroanalysis sponsible for recognition and repair of single-strand DNA-breaks, ofantiangiogenicactivity[8,9]. which might be induced by metal based drugs or oxidative stress PotentialofourRu(II)-p-cymenecomplexestomodulateseveral [15–17]. PARP also takes role in cell cycle progression and cell steps of metastatic progression like adhesion, migration, invasion, deathsignaling,andispotentialmoleculartargetfordevelopment proteoliticdegradationofextracellularmatrixandformationofnew of PARP inhibitors as chemo-sensitizer to DNA-damaging agents bloodvesselshasbeenstudiedinvitro. [18].InordertodeterminecontributionofPARPincellresponseto Intracellular distribution of ruthenium among protein and DNA ruthenium-basedcomplexesweinvestigatedwhetherPARPinhibi- cellularfractionsinHelacellswasanalyzedbyICP-OES,whileeffect tion by 3-aminobezamide may potentiate action of complex 5 on oncellcycleprogressionwasdeterminedbyFlowcytometry. HeLacells. Though thepharmacological target of rutheniumcomplexeshas notbeenunequivocallyidentified,itisgenerallyacceptedthattheir 2.Resultsanddiscussion cytotoxicityisrelatedtotheirabilitytobindDNA,comparablytoplat- inumbaseddrugs[10].CellDNA-damageresponsedependsuponthe 2.1.ResultsofMTTassay natureofthelesionandrecruitscomplexsystemofDNA-damagerec- ognition/repairproteins. AntiproliferativeactionofaseriesofRu(II)-arenecomplexesof Major mechanisms involved into repair of cisplatin induced structural formula [(η6-p-cymene)Ru(L1,2)Cl ], where L1,2 is 3- 2 DNA-lesions are reported to be Nucleotide Excision Repair (NER) acetylpyridine(1)and2-amino-5-chloropyridine(2),correspond- andMismatchRepair(MMR)mechanisms,androleofexcisionre- ingly, and [(η6-p-cymene)Ru(H L3,4)Cl], where H L3,4 represent 2 2 pair cross-complementing 1 (ERCC1) protein and MutS homolog 2,3-pyridinedicarboxylicacid(3),2,4-pyridinedicarboxylicacid(4) (MSH)proteinsincellresponsetoplatinumdrugshasbeenexten- havebeenreportedpreviouslyonsixtumorcelllinesshowingno sively investigated [11–14]. In this study we evaluated whether cytotoxicactivityupto200μM[6].Inthisstudyadditionalstructure N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 55 Table1 marginallythemostsensitivetothiscomplex.Prolongationofincu- ResultsofMTTassayforcomplexes1−5arepresentedasIC50values,obtainedafter48 bationtime(from48hto72h)didnotimprovetheantiproliferative or72htreatment. activity. Complex IC50(μM) Asalreadymentioned,thelowcytotoxicityinvitroisverycommon EA.hy926 MS1 HeLaa featureforruthenium-basedanticanceragentswhichwereshownina numberofcasestopossessgoodantitumor,andaboveall,antimetastatic 1 48h >300 >300 275.71±1.57 properties[20,21]. 72h >300 270.04±3.71 244.00±1.91 2 48h >300 220.82±15.98 >300 72h 258.46±2.79 218.91±6.21 283.20±3.12 2.2.Cellcycle 3 48h >300 >300 >300 72h 259.42±9.83 >300 >300 Theeffectofinvestigatedruthenium(II)-p-cymenecomplexeson 4 48h >300 >300 >300 72h >300 >300 >300 cellcycleprogressionofHeLaandEA.hy926cellswasexaminedby 5 48h 95.47±7.36 78.3±4.09 81.97±14.17 flow cytometry after continual treatment for 24h, using staining 72h 88.36±8.10 108.54±9.45 149.35±12.06 withPI[22].Examinationofthehistogramsindicatedthatonlycyto- a ResultsofactivityonHeLacellsarealreadypublished[6,7]. toxiccomplex5inducedperturbationsofcellcycleofHeLacells,and resultsarepresentedinFig.2.Othercomplexeshadnoeffectoncell cycle phase progression in both HeLa and EA.hy 926 (data not presented). activity comparison has been performed on endothelial cell lines Obtainedresultsshowthatcomplex5induceddecreaseofpercent (modelsystemsforinvitroangiogenesis)byMTTassay[19],includ- ofcellsinG1andslightarrestintheSphaseofcellcycle,withnoap- ingcompoundofstructuralformulae[(η6-p-cymene)RuCl(picolinic optotic effect (evaluated as Sub-G1 fraction). Slower progression acid)]·H O (5) [7]. Results are shown in Table 1. in terms of IC through replication phase indicates its preferential binding to DNA 2 50 values,determinedfrom3to4independentexperiments. andcorrelateswiththeresultsobtainedbyICP-OES(seebelow). Antiproliferativeactivityofinvestigatedcomplexesonendothelial cellswerelow,indicatedbyhighIC 50 values(>200μM),exceptfor 2.3.Intracellulardistributionofruthenium(II)inproteinvs.DNAfraction complex 5. Ruthenium compound 5 exhibited cytotoxic activity on allinvestigatedcelllinesafterboth48hand72hofcontinualincuba- Whileinvestigatingmechanismofactionofpotentialdrugcandi- tion,withIC 50 valuesfallinginthe80–150μMrange.Nosignificant datesitisimperativetolocatedruginteractionswithkeytargetsin tumor cell type selectivity was observed, although HeLa are the cell [20]. We separately analyzed ruthenium(II) distribution Fig.2.Effectofthecomplex5(IC50)oncellcycleprogressionofHeLacellsfollowing24hincubation.Histogramspresentedarerepresentativeofthreeindependentexperiments. amongtheDNAandproteinfractionsinHeLacellstreatedwithinves- Other rutheniumcomplexes have not shown any inhibitory effect in tigated ruthenium(II)-p-cymene complexes, using ICP-OES analysis, thisassay. after24htreatmentwith100μMofinvestigatedcomplexes.Eachin- vestigatedRu(II)compoundwasfoundintracellular,althoughexhi- 2.4.2.Invasionassay biting different levels of intracellular accumulation and different Inordertoinvestigateinvitropotentialofrutheniumcomplexes affinitiesforproteinorDNAbinding(Fig.3).Complex5waspreferen- toinhibitinvasionofHeLacells,weusedBDBioCoatMatrigelinva- tiallyaccumulatedinDNAfraction,showingdrasticallyhigheraffinity sion chambers as a system with the reconstituted basement mem- forDNA-bindingincomparisontootherinvestigatedRucomplexes, brane matrix-matrigel as an invasive substrate coated on a which may be a reason for its cytotoxic action. Level of ruthenium microporousfiltermembrane[26].Allinvestigatedrutheniumcom- bound to protein fraction was comparable among the tested com- plexes exhibited the potential of inhibiting HeLa cells invasion plexes.Complexes1,3,4exhibitednoDNAbinding,whichindicates (Fig.4A).Amongthenontoxicrutheniumcomplexeswhichwereap- thatligandstructuraldifferencesoftestedRu(II)compoundssignifi- pliedatequimolarconcentrations(100μM),complex2reducedinva- cantlyinfluencedaffinityorstabilityofDNAinteractions.Complexes sion with 33.6% of inhibition compared to the control (100% 1and2aswellas3and4havesomecharacteristicswhichprobably invasion).Cytotoxiccomplex5exhibitedinhibitoryactivityonHeLa obstructDNAinteractions,especiallydrugintercalationthroughη6- cell invasion up to 20%, although at lower concentration being bonded arene (such as rotation of the organic ligand about the 41μM(whichishalfofitsIC value). 50 Ru\N bond in complexes 1 and 2; and \COOH residue in meta- andpara-positionwithrespecttotheNatomofthearomaticringat 2.4.3.Adhesionassay complexes3and4). Pro-adhesive properties of investigated ruthenium(II)-p-cymene In contrast, complex 5 has a relatively planar surface, free from complexesonadhesionofHeLacellsonplasticsubstrate(uncoated substituents, which enables interactions with DNA thus inducing bottomoftheplates)wereevaluatedasresistancetotrypsintreat- theobservedantiproliferativeactivity. ment,whichisagoodmeasureofadhesivestrength[27].Wedeter- mined that after 90min treatment with 100μM concentration, 2.4.Invitroantimetastaticpotential complexes 2 and 3 contributed 3 times more to the adhesion of HeLacells,onplasticsurface,comparedtothatinthesametestper- Tumor metastasis is one of the major obstacles to cancer treat- formedwithcontrolcells(Fig.4B). mentandisacomplexmultistepprocessthatinvolvesinvasion,ad- Resultsoftheadhesionassayseemtobecompatiblewiththere- hesion, direct migration, and proteolitic activity to degraded ECM sults of the matrigel invasion assay (Fig. 4A). The same ruthenium barriers[23,24].Ru-basedcomplexesarereportedtoexhibitunique complexes2and3,whichamongthetestedcomplexescontributed antimetastaticpropertiesregardlessoftheircytotoxiceffect[20,25]. themosttothecelladhesion,alsoexhibitedthegreatestinhibitory Thus far, we investigated potential of our ruthenium(II)-p-cymene effect on invasion behavior of the cell, through Boyden chamber complexestoinhibitseveralstepsoftumorcellinvasionandangio- membranes. genesis, starting with analysis of their inhibitory activity on matrix metalloproteinases(MMPs),inHeLacells. 2.4.4.Inhibitionofcellmigration(Scratchassay) Woundmigrationassaywasusedasinvitroassaytodetermine 2.4.1.Gelatinzymography atfirstglancewhetherinvestigatedrutheniumcomplexeshavedi- MMPsare the family of tissue degrading enzymesthat facilitate rectinfluenceonmigrationofHeLacells[28].Forthisassayconflu- invasionandmetastasisthroughdissolutionofthebasementmem- entculturesofHeLacellswerewoundedandthenincubatedinfresh brane and degradation of the ECM. Among the MMPs, MMP-9 and completemedium.Migrationofcellswasobservedperiodicallyover MMP-2 (capable of degrading gelatin) are thought particularly im- timeunderthemicroscopeandrepresentativepicturesweretaken portantintheinvasionofthecancercells[24]. at 24h and 48h. Complex 5 induced inhibition of migration of TheactivitiesofsecretedMMP-2andMMP-9after24htreatmentof HeLa cells which was very visible after 48h incubation (Fig. 4D HeLaandEA.hy926cellswereanalyzedbygelatinzymography.Results shows results of wound assay for complex 5 only). Other tested show that only complex 5 exhibited inhibitory activity on secreted complexes1,2,3and4inducedmildinhibitioneffects(resultsnot MMP-2andMMP-9inHeLaandEA.hy926cells,respectively(Fig.4C). shown). 18 16 14 12 10 8 6 4 2 0 1 2 3 4 5 AND ro snietorp gµ/uR gp 56 N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 Proteins DNA Complexes Fig.3.Ruthenium(II)contentinHeLacells(DNAvs.proteinfractionofcells)after24htreatment,measuredbyICP-OES.Representativeexperiment. N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 57 Fig.4.Invitroantimetastaticpropertiesofthecomplexes.A.TheeffectofinvestigatedcomplexesonHeLacellinvasion,after24htreatmentand48hinvasiondeterminedbyMatrigel invasionassay.Dataarepresentedasthepercentagerelativetothecontrol(untreatedcells)andarerepresentativeofoneexperimentinduplicates.B.Theeffectofcompoundsonadhesion abilityofHeLacellstoplasticsurface(atadoseof100μMfor90min)determinebySRBassay.Dataarerepresentativeofoneoutofthreeseparateexperimentswithstandarddeviations, comparedtonontreatedcontrols.C.GelatinzymographyofHeLaandEA.hy926celltreatedfor24hwith100μMofinvestigatedcomplexes.D.Typicalresultsofmigrationassay(scratch assay)forHeLacellseitherintheabsenceorpresenceof0.5×IC50complex5.Centerofthegapisindicatedwiththespotline. 2.4.5.Tubeformationassay ruthenium complexes on angiogenesis in vitro, the tube formation Angiogenesisinvolvesendothelialcellproliferation,migrationand assaywasperformed.Whenseededontomatrigelinthepresenceofap- capillary tube formation. To examine the effect of our investigated propriatemedia MS1 (endothelial)cellsstarttoreorganize,showing Fig.5.Tubeformationassay(antiangiogenicpotentialinvitro).RepresentativemicrographsoftubeformationofMS1cellstreatedfor24hwithcomplex5(0.5×IC50)versus control. Fig.6.ExpressionlevelofERCC1mRNAdeterminedbyReal-Time-PCR(A)andproteinexpressionleveldeterminedbyWesternblot(B),inHeLacellsafter24hexposurewithinvestigated complexes2,3,4,5(100μM)andCDDP(7μM). projectionsandformingcell–cellcontactsandpolygonstructures[8].In DNAreplicationandhasbeenimplicatedintriggeringofapoptosisin- ourexperimentMS1endothelialcellsweretreatedwithsub-toxiccon- ducedbyanumberofcytotoxicanticanceragents[8].Basedonthe centrationsofinvestigatedcomplexestodeterminetheirpotentialto idea that MSH components mediate sensitivity to DNA damaging decreaseabilityofthesecellstoextendanddifferentiateintotube-like agents,weinvestigatedwhetherMSH2mightbeincludedincellre- structures.Ourcomplexes1,4,and5showedsomepotentialtoinhibit sponse to Ru(II)-p-cymene complexes, especially complex 5 which characteristicsofangiogenicbehaviorofendothelialcells.Complex5in- showed preferential affinity for DNA binding. We have analyzed ducedthehighestinhibitoryeffectamongthetestedcomplexes,andre- MSH2proteinexpressionlevelinHeLaafter24htreatmentwithin- sultsarepresentedinFig.5. vestigatedcomplexes,comparingtocisplatinandnontreatedcontrol. Data obtained indicated that protein level of MSH2 decreased after 2.5.EffectonexpressionofERCC1mRNAlevelbyQuantitativeReal-Time treatmentwithcomplex5,whilerestofthecomplexes,includingcis- PCR(RQ-PCR)andproteinlevel(Westernblot) platinasareference,hadnoeffect(Fig.7). Thereisconsiderableinterestintheinteractionsofruthenium(II) 2.7.Combinationaldrugstudy complexeswithDNAaswellasthebiologicalimpactoftheinteractions [10,29,30].Inordertoevaluatewhetherinvestigatedrutheniumcom- PARPisanuclearenzymeinvolvedinsinglestrand-breakrepair plexesinduceDNA-repair-dependentcellresponseastheresultofcyto- andmaybeinhibitedtoincreasechemosensitivitytowidespectrum toxicDNAlesions,weinvestigatedmRNAexpressionlevelofERCC1, of DNA-damaging agents including alkylating agents (cisplatin), andproteinexpressionlevelofERCC1andMSH2,whichmightbecoop- topoisomeraseIinhibitors(etoposide)etc.[17].Assumingthatruthe- erativelyinvolvedin cell responsetoDNA-bindinganticanceragents nium complex 5 bind to DNA, inducing lesions, we investigated [31].ERCC1iscriticalcomponentofNERwhichisprimarilyinducedin whether combination study of complex 5 with PARP-1 inhibitor 3- therepairofbulkyplatinum-DNAadducts,anditisknownthatitsre- aminobezamideonHeLacells,mayinterferewithcomplex5induced pairfunctionextendsbeyondNERandmayplayroleincellresponse cytotoxicity.Resultsofcombinationaldrugstudy(Fig.8)showedthat to the other type of DNA lesions [31]. Data obtained on HeLa cells 3-ABatconcentrationof2mMpotentiatedcytotoxicityofcomplex5 after24hcontinualtreatmentshowedthatcytotoxiccomplex5signif- which indicates that Ru complexes may induce DNA-single-strand icantlyincreasedexpressionofERCC1onmRNAlevel(Fig.6A).Other breaks,eitherastheresultofdirectbindingtoDNAorasanindirect rutheniumcompoundsshowednoeffectornegativelymodulatedrela- cytotoxiceffect[16]. tive expressionlevelsof ERCC1mRNA, comparedto thenon treated control, as determined by RQ-PCR. This data are in accordance with theresultsofWesternblot(Fig.6B),whichshowedincreasedprotein 3.Conclusion expressionlevelinducedbycomplex5,incomparisontothecontrol. ERCC1has been investigatedasa markerof cisplatin resistance, and Based on their in vitro anticancer properties, investigated com- ERCC1mRNAexpressioncorrelatestosensitivitytocisplatintosome plexes have one complicated mechanism of action, involving both extend[12],thusfarcisplatinhasbeenusedasreferencecompound in this study. After 24h cisplatin treatment in HeLa cells, ERCC1 120 mRNAexpressionwasslightlydownregulated,whichwasoppositeto theeffectofRucomplex5.Anadditionalstudyisnecessarytoaddress 100 the role of ERCC1 upregulation in response to complex 5 induced cytotoxicity. 80 2.6.EffectonexpressionofMSH2proteinlevel(Westernblot) 60 MSH2isacomponentofmismatchrepairsystem(MMR)which 40 has central role in correcting mismatches in DNA occurring during 20 0 0 0.5xIC50 IC50 1.5xIC50 Fig.7.MSH2proteinexpressionlevelinHeLacellsafter24htreatmentwithcomplexes 1–5(100μM)andCDDP(7μM)determinedbywesternblot. )lortnoc fo %( htworg lleC 58 N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 Complex 5 Complex 5+1mM 3-AB Complex 5+2mM 3-AB Fig.8.CellgrowthinhibitiongraphofHeLacellsafter48hcontinualactionofcombined treatmentofcomplex5(atconcentrationscorrespondingto0.5×IC50,IC50and1.5xIC50) and3-AB(atconcentrations1mMand2mM),determinedbySRBassay.Dataarerepre- sentativeforoneoutofthreeseparateexperimentswithstandarddeviations. N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 59 extraandintra-cellularprocesses,quitedistinctfromthatofclassical weremaintainedasmonolayercultureinnutrientmedium,Dulbecco's platinum-basedcompounds. ModifiedEagleMedium(DMEM).PowderedDMEMmediumwerepur- In accordance with the previously published results [6,7] where chasedfromSigma-AldrichCo.NutrientmediumDMEMwaspreparedin replacement of pyridine ligands in the molecule of the ineffective sterileionizedwater,supplementedwithpenicillin(192IU/ml),strepto- ruthenium(II)-arene (complexes 1–4), with picolinic acid, resulted mycin(200µg/ml),4-(2-hydroxyethyl)piperazine-1-ethanesulfonicacid in an enhancement of antiproliferative activity, additional study on (HEPES) (25mM), L-glutamine (3mM) and 10% of heat-inactivated twoendothelialcelllines(EA.hy926,MS1)showedagainthatpicoli- fetalcalfserum(FCS)(pH7.2)andD-glucose(4.5g/l). nate ruthenium(II)–arene (complex 5) exhibited enhanced activity, The cells were grown at 37°C in 5% CO and humidified air 2 comparedtotheothercomplexes.Complex5exerteditscytotoxicef- atmosphere. fect through S phase cell cycle arrest and significant preference for DNAbinding.Ourinvestigationofintracellulardistributionofruthe- 4.2.2.MTTassay niumcomplexesallowedustoconcludethatDNA,ratherthanpro- Antiproliferativeactivityofcomplexes1–5wasdeterminedusing teins,maypresenttargetforcomplex5cytotoxicaction.Moreover, 3-(4,5-dymethylthiazol-yl)-2,5-diphenyltetrazolium bromide (MTT, it upregulated expression of DNA repair enzyme ERCC1 on both Sigma-AldrichCo.)assay[19].Cellswereseededin96-wellcellcul- mRNAandproteinlevel,andaffectedproteinlevelofMSH2,which tureplates(NUNC)EA.hy926(4000c/w)andMS1(7000c/w)incul- indicated involvement of DNA-repair-dependent cell response. Po- ture medium and grown for 24h. Compounds were dissolved in tentiationofactionofcomplex5incombinationwithPARP-1inhibi- DMSOandfurtherdilutedwithnutrientmediumtodesiredconcen- tor3-AB,suggestsimportanceofunderstandingcellularpathwaysor trations(upto300μM).FinalconcentrationofDMSOneverexceeded targetsutilizedbyrutheniumcomplexeswhichshowenhancedsen- 1%.Solutionsofvariousconcentrationsofexaminedcompoundwere sitizationoftumorcellsincombinationwithPARPinhibition. addedtothewells,exceptthecontrolwellswhereonlynutrientme- Complexes1–4whichpreferentiallyaccumulatedinproteinfrac- diumwasadded.Alltestswerecarriedoutintriplicate.Afterincuba- tionhadminorgrowthinhibitoryeffectandmodestanti-invasivepo- tion periods of 48 and 72h 20μL of MTT solution, 5mg/mL in tential.Complex5showedinhibitionofMMP-sactivity,inhibitionof phosphatebuffersolution,pH7.2wereaddedtoeachwell.Samples migrationofHeLacells,andinhibitionoftubeformationofMS1endo- were incubated for 4h at 37°C, with 5% CO in humidified atmo- 2 thelialcells(antiangiogenicpotentialinvitro).ComparablytoNAMI- sphere. Formazan crystals were dissolved in 100µl of 10% sodium Alikerutheniumcompounds(3),Ru(II)complex5showedcapacity dodecyl sulfate (SDS). Absorbances were recorded on an enzyme- tomodifyimportantparametersofmetastasis.However,thenature linked immunosorbent assay (ELISA) reader (ThermoLabsystems of the target(s) responsible for the antimetastatic activity remains Multiskan EX 200–240V) after 24h at the wavelength of 570nm. tobeelucidated. IC (μM)valuesweredeterminedfromthecellsurvivaldiagrams. 50 Structure–activity comparison revealed heterogenic mode of ac- tivity of tested ruthenium(II) complexes with regard to antimeta- 4.3.Flow-cytometricanalysisofcellcyclephasedistribution static effect and cytotoxic potential. Variations of the nature of monodentateorbidentateligandsignificantlyinfluencedpharmaco- Quantitative analysis of cell cycle phase distribution was per- logicalpropertiesofthecomplexes.Moreover,areneidentitymayaf- formed by flow-cytometric analysis of the DNA content in fixed fect drug-uptake and interactions with the potential intracellular HeLaandEA.hy926cells,afterstainingwithpropidiumiodide(PI) targetssuchasDNA,contributingtononcovalent,hydrophobicinter- [22]. actionswithDNAbases,suchasintercalation(29). Cellswereseededat densityof2×105cells/wellat6-wellplate Basedonthedataobtainedinthisresearch,picolinaterutheniu- andgrowthinnutritionmedium.After24hcellswerecontinuallyex- m(II)–arene(complex5)distinguishedamongthetestedruthenium posedtoinvestigatedcompounds1–4(300μM)and5(82μM/IC ). 50 complexes,according toitshighestantiproliferativeandantimetas- After24hofcontinualtreatmentcellswerecollectedbytrypsiniza- taicpotentialinvitro.Replacementofsubstitutedpyridineligand(in tion, washed twice with ice-cold PBS, and fixed for 30min in 70% complexes1,2,3,4)withligandofplanarsurface,freefromsubstit- EtOH.AfterfixationcellswerewashedagainwithPBS,andincubated uentssuchaspicolinicacid(complex5)allowedforbetterDNAbind- with RNaseA (1mg/mL) for 30min at 37°C. Just before flow- ingaffinityandcytotoxicity. cytometric analysis cells were stained with PI at concentration of DuetotheinteractionwithDNAandinvitroantimetastaticchar- 400μg/mL.Cellcyclephasedistributionwasanalyzedusingafluores- acteristicswebelievethatcomplex5ispromisingcandidateforfur- cence activated sorting cells (FASC) Calibur Becton Dickinson flow therinvivoantitumorstudy. cytometerandCellQuestcomputersoftware. 4.Experimentalprotocols 4.4.Intracellulardistributionofinvestigatedcomplexes(proteinsvs.DNA fraction) 4.1.Synthesis Ruthenium level in proteins vs. DNA fraction after treatment of The complexes of structural formula [(η6-p-cymene)Ru(L1,2)Cl ], HeLa cells with investigated complexes was determined using ICP- 2 whereL1,2is3-acetylpyridine(1)and2-amino-5-chloropyridine(2), OES. 6×106cells/75cm2 falcon dish (NUNC) were treated with in- correspondingly,and[(η6-p-cymene)Ru(H L3,4)Cl],whereH L3,4rep- vestigatedcomplexesin100μMconcentrationsfor24h.Aftertreat- 2 2 resent2,3-pyridinedicarboxylicacid(3),2,4-pyridinedicarboxylicacid mentcellswerewashedbyicecoldPBSandharvestedbyscraping. (4),and[(η6-p-cymene)RuCl(picolinicacid)]∙H O(5)weresynthesized Aftercentrifugationat2000rpm,10minat4°Ccellpelletwasfrozen 2 andcharacterizedaspreviouslydescribed[6,7]. at−70°C. DNA and proteins were isolated using TRI Reagent® (Sigma- 4.2.AnalysisofcellgrowthinhibitionbyRu(II)-complexes AldrichCo.)accordingtomanufacturer'sprocedureandconcentrations determined spectrofotometrically by measuring absorbance at A260 4.2.1.Cellculture andA280nmrespectively(EppendorfBioPhotometer6131).Rutheni- TheendothelialcelllinesEA.hy926(permanenthumancelllinede- umlevelsweredeterminedinisolatedDNAandproteinfraction,after rived from a fusion of human umbilical vein endothelial cells-HUVEC sample preparation for ICP-OES according to standard procedure, andhumanlungadenocarcinomacells-A549)andMS1(murineendo- usingThermoScientificiCAP6500DuoICP(ThermoFisherScientific, thelialcellsimmortalizedbyinfectionencodingSV40largeTantigen) Cambridge,UK) 60 N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 4.5.Invitroantimetastaticpotential andabsorbancesrecordedonanenzyme-linkedimmunosorbentassay (ELISA) reader (ThermoLabsystems Multiskan EX 200–240V) at the 4.5.1.Gelatinzymography wavelengthof570nm.Eachexperimentwasdoneintriplicate. Gelatinolitic activity of secreted MMP-2 and MMP-9 in drug- treated HeLa and EA.hy 926 cells was analyzed by zymography in 4.5.4.Cellmigrationassay(Scratchassay) 10%SDS-polyacrylamidegelsimpregnatedwith0.1%gelatin. Togainfirstinsightsinthepotentialofinvestigatedrutheniumcom- Cellswereseededin6-wellplate(NUNC)induplicate(6×105c/w) plexestoinhibitmigrationofHeLacells,invitroscratchassaywasdone in nutrition mediumbeing RPMIfor HeLaand DMEM forEA.hy 926 [22].4×105cells/wellwereseededinto6-wellcellcultureplateandcul- cells.At70%ofconfluencecellswereexposedtoinvestigatedcomplexes turedinRPMI1640with10%FCSintissueculture5%CO incubatorto 2 (100μM)inserum-freemedium,for24h.Equalamountsofsuperna- confluence. A scrape was made in the center of the cell monolayer tantproteinsforeachsample,asdeterminedbyBradfordassay[32], withasterilepipettetip(p200)tocreateagapofconstantwidth.Cells weremixedwithloadingbuffer(loadingbuffer:0.125MTris–HCl,pH detachedbyscrapingwereremovedbygentlywashingwithPBS.The 6.8;20%glycerol;10%SDSand0.25%bromophenolblue)beforeloading initialimagesofthewoundswerecapturedwithOlympusdigitalcam- to the SDS-polyacrylamide gels. Electrophoresis was performed at eraconnectedtotheinvertedmicroscope(CarlZeiss,Jena,Germany,ob- 125V, gels were washed 3 times for 20min in renaturing solution jective6.3/0.20)andcellswerefurtherincubatedfor24hand48hin (2.5% Triton X-100) followed by a 48h incubation at 37°C in the fresh medium for control or with investigated complexes in desired enzyme assay buffer (50mM Tris–HCl, pH 7.5; 0.2M NaCl; 5mM concentrations1–4(100μM)and5(0.5×IC ).Picturesweretakento 50 CaCl ; 0.05% NaN ) for the development of enzyme activity bands. qualitativelydeterminetheinfluenceofinvestigatedcomplexesonmi- 2 3 Afterincubation,thegelswerestainedwith0.05%Coomassiebrilliant gration.Photographsoftheinitialwoundedmonolayerswerecompared blueG-250inamixtureofMeOH:AcOH:H O(2.5:1:6.5)anddestained withthecorrespondingimagesoftreatedcells.Artificiallinetraversing 2 in 4% MeOH with 8% AcOH. In order to increase the sensitivity, the throughthemiddleofgapwasdrawnonpicturesoforiginalwounds gelswerefurtherdestainedinasolutionof1%TritonX-100(1–2h) andoverlaidontheimagesofculturesafterincubation. [24,33].Thegelatinolyticactivitieswerevisualizedascleartransparent bandsagainstthebluebackgroundofCoomassiebrilliantblue-stained 4.5.5.Tubeformationassay(invitroangiogenesisassay) gelatin. Potentialofinvestigatedcomplexestoinhibitangiogenesisinvitro was analyzed by tube formation assay in mouse endothelial cells 4.5.2.Matrigelinvasionassay (MS1). MS1 cells plated into gel of basement membrane proteins, Thepotentialofourruthenium(II)-p-cymenecomplexes,toinhib- rapidlyorganizeintomulticellulartube-likestructures,whileantian- itinvasionofHeLacellswecarriedoutbymatrigelinvasionassayin giogeniceffectoftestedcompoundsisobservedasreductionoftube 24-well BD BioCoat Matrigel invasion chambers (BD Biosciences) formation[8].Briefly,24-wellplateswerecoatedwithcollagenand [26].2×105cells/wellofHeLacellswereseededin6-wellcellculture allowed to solidify at 37°C for 1h. MS1 (mouse endothelial cells) plate(NUNC)innutritionmedium.After24hcellsweretreatedwith wereseeded1×105cells/wellinmedium.Treatmentwasadded2h investigated ruthenium complexes 1–4 (100μM) and 5 (0.5×IC ) after cells settled (for complexes 1–4 concentration was 100μM, 50 for24h.Controlwerecellsinthenutritionmediumonly.Aftertreat- and for complex 5 concentration was 0.5×IC ). Tube formation 50 mentcellswereleftfor4hinserum-freeRPMI1640mediumwith wasobservedperiodicallyovertimeundermicroscopeandrepresen- 0.2%BSA.Afterstarvationcellswereharvestedbytrypsinizationand tativepicturesweretakenafter24hincubationwithOlympusdigital resuspended in serum-free RPMI 1640 medium with 0.2% BSA. cameraconnectedtotheinvertedmicroscope(CarlZeiss,Jena,Germany, 4×104cells/0.5mL/wellwereseededontothematrigel-coatedfilters objective6.3/0.20). (with8μm poresize)inBioCoatMatrigelinvasionchambers.RPMI 1640 medium with 10% FCS was added to the lower compartment 4.6.QuantitativeReal-TimePCR(RQ-PCR) aschemoattractant.Afterincubationinahumidifiedtissueculturein- cubator,at37°C,5%CO atmospherefor48h,thenon-invadingcells 4.6.1.RNAisolationandcDNAsynthesis 2 areremovedfromtheuppersurfaceofthemembraneby“scrubbing”. RNAwasisolatedfrom6×106HeLacells,aftertreatmentwithin- Cellsthatinvadedthelowersurfaceofthefiltersweredetachedand vestigated complexes 2–5 (100μM) for 24h, using TRI Reagent® stainedwithdetachmentbufferTrypsine/EDTA(0.05%/0.02%)inPBS (Sigma-Aldrich) according to the manufacturer's procedure. Total with5μMCalcein-AM.200μLoftheTrypsine/EDTAsolutioncontain- RNAofeachsamplewasquantifiedbyaspectrophotometer(Eppen- ingthedislodgedcellswastransferredfromeachwellof24wellplate dorfBioPhotometer)measuringtheabsorbanceat260nm.Theratio toa96-wellplateinduplicate.Invadedcellswerequantifiedin96- A /A wasusedtoassessthepurityofisolatedRNA.TheRNAcon- 260 280 well plate with the fluorescence microplate reader (Fluoroskan As- centrationwascalculatedinμg/μL.ToanalyzetheRNAbandingpat- cent FL, Thermo Electron Corporation) at an excitation wavelength tern, gel electrophoresis (Pharmacia Biotech) in 2% agarose gels of485nmandanemissionwavelengthof520nm.Assaywasdone with ethidium bromide was performed. The RNA samples were induplicate. mixedwithxylenecyanolcolor(0.25%xylenecyanolin30%glycerol inwater)andrunin0.5×TBEbufferfor50minat25mA.TheRNA 4.5.3.Celladhesionassay(resistancetodetachmentassay) bandswerevisualizedonaUVtransilluminator(Hoefer).Toperform To evaluate potential influence of investigated ruthenium(II)-p- RT-PCRwithrandomprimers,2μgoftotalRNAwasusedastemplate cymenecomplexesonadhesionoftumorcells,theadhesionofHeLa for MultiScribe™ Reverse Transcriptase (50U/μL) in High-Capacity cellsonplasticsubstrate(uncoatedbottomoftheplates)wasevalu- cDNA Reverse Transcription kit (Applied Biosystems), according to atedaccordingtomethodreportedby[27]. themanufacturer'sinstructions.Thereactionmixtureswereincubat- HeLacellsweregrownatadensityof3×104cells/mL,in0.2mL/ edinMastercyclergradient(Eppendorf). wellnutritionmediumina96-wellplasticplate(NUNC)andincubated at37°C.At70%confluencecellsweretreatedwithinvestigatedcom- 4.6.2.QuantitativeReal-TimePCR(RQ-PCR) plexes at 100μM concentrations, for 90min.Briefly,30μL of(0.05%, TheanalysisofgeneexpressionlevelofERCC1andGAPDH(endoge- w/v) trypsin solution per well was used to detach cell at 37°C for nous control) was done by using TaqMan® Gene Expression Assay 30min.Nondetachedcells were fixed with 200μl of10% (w/v) cold (ERCC1probe,AppliedBiosystem,ID:Hs_01012155_g1,andTaqMan® trichloroacetic acid (TCA) at 4°C for 1h. Number of adherent cells MGBGAPDHprobe,AppliedBiosystem,ID:Hs_00355782_m1),onABI was determined using sulforhodamine B (SRB; Sigma) staining [30], PRISM® 7500 PCR instrument (Applied Biosystems Inc., Foster City, N.Gligorijevićetal./JournalofInorganicBiochemistry108(2012)53–61 61 CA).PCRreactionswerepreparedinduplicateforeverysamplein96- HEPES 4-(2-hydroxyethyl)piperazine-1-ethanesulfonicacid well optical plates. Each well contained 11μL of TaqMan® Universal FCS Fetalcalfserum PCR Master Mix (Applied Biosystems), 1μL of cDNA (100ng/μL)and MTT 3-(4,5-dymethylthiazol-yl)-2,5-diphenyltetrazolium 8μLH O.Notemplatecontrolinduplicateaswell,wasusedasaPCR bromide 2 negativecontrol(11μLTaqMan®UniversalPCRMasterMix(Applied ELISA Enzyme-LinkedImmunosorbentAssay Biosystems),9μLH O).PCRprofileconsistedofthefollowingsteps:en- FASC FluorescenceActivatedSortingCells 2 zymeactivation50°C2min;initialdenaturation95°C10min;40cy- PBS PhosphateBufferedSaline cles of denaturation at 95°C 15s, and annealing/extension at 60°C TCA Trichloroaceticacid 1min. SRB SulforhodamineB CDDP Cis-diamminedichloroplatinum(II) 4.7.Westernblot ECL EnhancedChemiluminescence GAPDH Glyceraldehyde3-phosphatedehydrogenase HeLa cells were treated with investigated ruthenium complexes 1–5 in concentration 100μM and CDDP in concentration 7μM for 24h. Control were untreated cells, maintained in medium only. After treatment cells were washed with phosphate buffered saline Acknowledgments (PSB)(Sigma-AldrichCo.)andwholecelllysateswereharvestedin 100μLofLysisbuffercontaining:37mMTris–HCl,75mMNaCl,25% ThisworkwassupportedbytheMinistryofScience,Republicof TritonX-100,and10μlproteaseinhibitorcocktail(Sigma).Samples Serbia,Grant,No.III41026andGrant,No172017. werestoredat−20°Cuntilfurtheranalysis. Proteinconcentrationsoftheresultinglysatesweredetermined References using Bradford assay [32]. Whole cell lysates were mixed with Laemmliesloadingbufferandsubjectedfor1htoSDS-PAGE(10% [1] E.S.Antonarakis,A.Emandi,CancerChemother.Pharmacol.66(2010)1–9. PAA gel) at 120V, followed by electroblotting to nitrocellulose [2] M.Galanski,V.B.Arion,M.A.Jakupec,B.K.Keppler,Curr.Pharm.Des.9(2003) membrane(0.45μm)(Bio-RadLaboratories,Inc.)for1hat100V. 2078–2089. [3] A.Vacca,M.Bruno,A.Boccarelli,M.Coluccia,D.Ribatti,A.Bergamo,S.Garbisa,L. Membranes were blocked for 1h with 5% non-fat milk in PBS at Sartor,G.Sava,Br.J.Cancer86(2002)993–998. roomtemperatureandsubsequentlyprobedovernightwithapuri- [4] S.Arandjelovic,Z.Tesic,P.Perego,L.Gatti,N.Carenini,F.Zunino,R.Leone,P. fied mouse anti-human ERCC1 monoclonal antibody (1:500 dilu- Apostoli,S.Radulovic,Med.Chem.2(2006)227–237. [5] A.Levina,A.Mitra,P.A.Lay,Metallomics1(2009)458–470. tion)(BDBiosciencesPharmingen),orpurifiedmouseanti-human [6] S.Grgurić-Šipka,I.Ivanović,G.Rakić,N.Todorović,N.Gligorijević,S.Radulović, MSH2 monoclonal antibody (1:1000 dilution) (BD Biosciences V.B.Arion,B.K.Keppler,Ž.Lj.Tešić,Eur.J.Med.Chem.45(2010)1051–1058. Pharmingen).AfterwashingthemembraneswithTris-bufferedsa- [7] I.Ivanović,S.Grgurić-Šipka,N.Gligorijević,S.Radulović,A.Roller,Ž.Lj.Tešić,B.K. Keppler,J.Serb.Chem.Soc.76(2011)53–61. linethreetimes,themembraneswereincubatedwithanti-mouse [8] E.Aranda,G.I.Owen,Biol.Res.42(2009)337–389. IgG-peroxidase conjugated secondary antibody (1:2000 dilution) [9] J.L.Arbiser,etal.,Proc.Natl.Acad.Sci.94(1997)861–866. (Sigma-Aldrich Co.) for 1h. The levels of specific proteins were [10] V.Brabec,O.Novakova,DrugResist.Updates9(2006)111–122. [11] C.Lanzi,P.Perego,R.Supino,S.Romanelli,T.Pensa,N.Carenini,I.Viano,D.Colangelo, detectedusingenhancedchemiluminescence(ECL)detectionsys- R.Leone,P.Apostoli,G.Cassinelli,R.A.Gambetta,F.Zunino,Biochem.Pharmacol.55 tem(AmershamBiosciences),followedbyautoradiography. (1998)1247–1254. [12] Y.Tomoda,M.Katsurai,M.Okajima,N.Hosoya,N.Kolmo,K.Miyagama,Int.J. Cancer124(2009)2997–3001. 4.8.Combinationaldrugstudy [13] L.P.Martin,T.C.Hamilton,R.J.Schilder,Clin.CancerRes.14(2008)1291–1295. [14] E.Reed,Clin.CancerRes.11(2005)6100–6102. Inordertodeterminethesensitivityoftumorcellstoruthenium [15] D.D'amours,S.Desnoyers,I.D'silva,G.G.Poirier,Biochem.J.342(1999)249–268. complex 5 in combination with PARP inhibitor 3-aminobenzamide, [16] S.Kapitza,M.A.Jakupec,M.Uhl,B.K.Keppler,B.Marian,CancerLett.226(2005) 115–121. sulforhodamine B (SRB) test was done [34]. Briefly, 2000 cells per [17] T.Helleday,E.Petermann,C.Lundin,B.Hodgson,R.A.Sharma,Nat.Rev.Cancer wellwereseededin96-wellplates,anddayaftercellsweretreated 8(2008)193–204. by ruthenium compound 5 alone or in combination with 3-AB for [18] V.Schreiber,F.Dantzer,J.C.Ame,G.deMurcia,Nat.Rev.Mol.CellBiol.7(2006) 517–528. 48hofcontinualaction.Absorbanceswererecordedonanenzyme- [19] R.Supino,MethodsinMolecularBiology,in:S.O'Hare,C.K.Atterwill(Eds.),Vitro linked immunosorbent assay (ELISA) reader (ThermoLabsystems ToxicityTestingProtocols,HumanaPress,NewJersey,1995,pp.137–149. MultiskanEX200–240V)atthewavelengthof570nm.Eachexperi- [20] P.J.Dayson,G.Sava,DaltonTrans.16(2006)1929–1933. [21] I.Kostova,Curr.Med.Chem.13(2006)1085–1107. mentwasdoneintriplicate. [22] M.G.Ormerod,AnalysisofDNA—GeneralMethods,in:M.G.Ormerod(Ed.),Flow Cytometry, a Practical Approach, Oxford University Press, New York, 1994, Abbreviations pp.119–125. [23] L.A.Liotta,CancerRes.46(1986)1–7. [24] K.Shravan,A.O.Francis,M.Sanjeera,Clin.Exp.Metastasis15(1997)361–367. [25] S.Aranđelović,S.Bjelogrlić,N.Malešević,Ž.L.J.Tešić,S.Radulović,J.BUON14 (2009)271–279. [26] A.Albini,Y.Imamoto,K.Kleinman,G.R.Martin,S.A.Aronson,J.M.Kozlowski,R.N. ECM Extracellularmatrix McEwan,CancerRes.47(1987)3239–3245. ICP-OES InductivelyCoupledPlasma-OpticalEmissionSpectrometry [27] G.Sava,F.Fransin,M.Cocchietto,F.Vita,E.Podda,P.Spessotto,A.Furlani,V.Scarcia, NER NucleotideExcisionRepair G.Zabucchi,Eur.J.Cancer40(2004)1383–1396. MMR MismatchRepair [28] C.C.Liang,A.Y.Park,J.L.Guan,Nat.Protoc.2(2007)329–333. [29] O.Novakova,H.Chen,O.Vrana,A.Rodger,P.J.Sadler,V.Brabec,Biochemistry42 ERCC1 ExcisionRepairCross-Complementing1 (2003)11544–11554. MSH2 MutShomolog2 [30] A.Pizzaro,P.J.Sadler,Biochimie91(2009)1198–1211. PARP Poly(ADP-ribose)polymerase [31] H.Zhenga,X.Wanga,R.J.Legerskic, P.M.Glazerd, L.Li,DNARepair5(2006) 566–574. 3-AB 3-aminobezamide [32] M.Bradford,Anal.Biochem.72(1976)248–254. PI Propidiumiodide [33] C.Fernandez-Patron,C.Zouki,R.Whittal,J.S.D.Chan,S.T.Davidge,J.G.Filep,Biol. MMP Matrixmetalloproteinases Proced.Online4(2002)38–48. [34] P.Skehan,R.Storeng,D.Scudiero,A.Monks,J.McMahon,D.Vistica,J.T.Warren, RPMI RoswellParkMemorialInstitute H.Bokesch,S.Kenney,M.R.Boyd,J.Natl.CancerInst.82(1990)1107–1112. DMEM Dulbecco'sModifiedEagleMedium