Structure-activity relationships and DNA binding properties of apoptosis inducing cytotoxic rhodium(III) polypyridyl complexes containing the cyclic thioether [9]aneS(3).

JournalofInorganicBiochemistry103(2009)698–708 ContentslistsavailableatScienceDirect Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio Structure–activity relationships and DNA binding properties of apoptosis inducing cytotoxic rhodium(III) polypyridyl complexes containing the cyclic thioether [9]aneS 3 Ruth Biedaa, Ingo Ottb, Mara Dobroschkec, Aram Prokopc, Ronald Gustb, William S. Sheldricka,* aLehrstuhlfürAnalytischeChemie,Ruhr-UniversitätBochum,Universitätsstraße150,D-44780Bochum,Germany bInstitutfürPharmazie,FreieUniversitätBerlin,Königin-Luise-Straße2+4,D-14195Berlin,Germany cDepartmentofPediatricOncology/Hematology,UniversityMedicalCenterCharitéBerlin,13353Berlin,Germany a r t i c l e i n f o a b s t r a c t Articlehistory: TheRh(III)polypyridylcomplexesofthetype[RhCl(pp)([9]aneS )]2+[(pp)=2,20-bipyridine(bpy),2,20- 3 Received24September2008 bipyrimidine (bpm),1,10-phenanthroline (phen), pyrazino[2,3-f]quinoxaline (tap), dipyrido[3,2-d:20,30- Receivedinrevisedform5January2009 f]quinoxaline (dpq), dipyrido[2,3-a:20,30-c]phenazine (dppz)] 2–7 have been prepared in a stepwise Accepted7January2009 Availableonline22January2009 mannerbytreatmentofRhCl 3(cid:2)3H 2 Owiththeappropriatepolypyridylligand(pp)followedby1,4,7-tri- thiacyclononane.InteractionsofthepolypyridylcomplexeswithDNAwereinvestigatedbyCDandUV/ visiblespectroscopyandbygelelectrophoresis.Thedpqcomplex6cleavesDNAexiguouslyinthedark, Keywords: butUVirradiationisrequiredtoinducenucleaseactivityforthebpycomplex2.Whereas2[IC values: Rhodium 50 12.8(±0.2)and4.4(±0.1)lM]exhibitssignificantlyhighercytotoxicitiestowardsMCF-7andHT-29cells Polypyridylligands than4[IC values:36.3(±6.0)and72.2(±8.0)],theactivityofcomplexesintheseries4/6/7correlates Thioethercomplexes 50 DNAcleavage directlywiththesizeofthepolypyridylligand,asdocumentedbytheirrespectiveIC 50 valuesof72.2 Antitumoragents (±8.0),20.9(±2.8)and7.4(±2.2)towardsHT-29cells.Complexesofthenitrogen-richligandsbpm(3) [IC values:1.7(±0.5)and1.9(±0.1)lM]andtap(5)[IC values:11.5(±0.6)and7.6(±4.8)lM]arecon- 50 50 siderablymorepotentthantheirbpyandphencounterparts2and4.Measurementofthelactatedehy- drogenasereleaseforlymphoma(BJAB)cellsafter1hincubationdemonstratesthatunspecificnecrosisis negligibleforthemostactivecompounds3and7.SpecificcelldeathapoptosisviaDNAfragmentation wasdetectedforBJABcellsafter72hincubationandsignificantlossofthemitochondrialmembrane potentialinlymphomacellsindicatesthattheintrinsicpathwayisinvolved. (cid:2)2009ElsevierInc.Allrightsreserved. 1.Introduction mentmembrane[8].Astudyonthecytotoxicityof[RuCl(en)([9]a- neS )](CF SO ) has recently been published [9] but low activity 3 3 3 Althoughthereisgreatcurrentinterestinthedevelopmentof wasreportedwithIC valuesofonly65lMand175lMbeingob- 50 nonplatinum metal complexes for tumor therapy [1–3], studies servedfortheTS/AandHBL-100celllinesafteranincubationper- on the cytotoxicity of group 9 transition metal complexes have iodof72h. beenratherlimitedandhaveconcentratedondirhodium(II,II)car- Systematic studies on the polypyridyl (pp) compounds mer- boxylates [4–6]. Very recent work on the in vitro cytotoxicity of [RhCl (DMSO)(pp)] and [g5-C Me )RhCl(pp)](CF SO ) (pp)=bpy, 3 5 5 3 3 rhodium(III)compoundshas,however,producedsomepromising phen,dpq,dppz,dppn)havedemonstratedthattheinvitrocyto- results.Forinstancemer-[RhCl (tpy)]isactivetowardsHCV29Ttu- toxicitiesofthesecontrastingtypesofrhodium(III)complexesto- 3 morcells(ID =6.0lM)[7]andasimilarlevelofcelltoxicityhas wards the cancer cell lines HT-29 und MCF-7 increase with 50 beenestablishedforfac-[RhCl ([9]aneNS )][8].Cytotoxicityassays increasing polypyridyl ligand size [10,11]. The meridional com- 3 2 demonstratethatfac-[RhCl ([9]aneNS )]isasactiveascisplatinto- plexes mer-[RhCl (DMSO)(pp)] (pp=dpq, dppz, dppn) are extre- 3 2 3 wardstheovariancancercelllineNuTu-19.Inaddition,cellinva- melypotentandexhibitIC valuesintherange0.051–0.095lM 50 sion assays against NuTu-19 and Ovepi cell lines show that the towards these human cancer cell lines. Facial iridium(III) com- facial rhodium(III) complex is more efficient than cisplatin at plexesofthetypefac-[IrCl (DMSO)(pp)]areconsiderablylessac- 3 increasing the invasion of the cancerous cells through the base- tive with IC values between 6.1 and 0.21lM for the same 50 polypyridylligands[12]. Polypyridyl complexes that can recognize and bind at specific * Correspondingauthor.Tel.:+492343224192;fax:+492343214420. E-mailaddress:william.sheldrick@rub.de(W.S.Sheldrick). DNA sites have received considerable attention [13,14]. Whereas 0162-0134/$-seefrontmatter(cid:2)2009ElsevierInc.Allrightsreserved. doi:10.1016/j.jinorgbio.2009.01.008 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 699 octahedraltransitionmetalcomplexescontainingthreebipyridine 3 orphenanthrolineligandshavebeenshowntobegroovebinders Cl 2 4 or possible partial intercalators [13,15], strong intercalative binding has often been established when a larger polypyridyl li- Cl N 5 Rh gand such as phi or dppz is present. Head-on intercalation has, H for instance, been confirmed for the phi ligand of D-a-[Rh{(R,R)- O N 6 Me trien}(phi)]3+inthecrystalstructureofitsadductwithahexa- Cl nuc 2 leotide[16].Theimportantroleofancillaryligandsinenabling 9 7 H C 8 specificDNAsitebindingisdemonstratedbytheHdonorligands 3 en and [12]aneN in [Rh(en) (phi)]3+ and [Rh([12]aneN )(phi)]3+, 1 4 2 4 whichcanrecognize50-GC-30 sequences,whereastheirstructural analogue [Rh([12]aneS )(phi)]3+ with the H acceptor thioether Scheme2. Structureofcomplex1. 4 coligand[12]ane-S doesnot[14,17].Incontrasttothesephicom- 4 plexes,thealternativeside-onintercalationmodehasbeenestab- lishedforthedppzligandinhalf-sandwichcomplexesofthetypes [(g5-C Me )Ir(L)(dppz)]2+ and [(g6-C Me )Ru(L)(dppz)]2+ with sizeandhydrophobicityofthepolypyridylligandand(c)thepres- 5 5 6 6 thioether-coordinated methionine-containing amino acids or enceofHacceptoratomsinthesearomaticligands.Therestricted peptides [18,19]. It is interesting to note that both [(g5- abilityofcyclicthioetherstoneutralizepositivechargethroughr C Me )IrCl(dppz)]+ [20] and [g5-C Me )RhCl(dppz)]+ [11] exhibit donation[26,24]willleadtoadrasticdecreaseintherateofCl(cid:3)/ 5 5 5 5 significantinvitrocytotoxicitytowardsHT-29andMCF-7cellsin H Oexchangeincomparisontotheanalogous[g5-C Me ](cid:3)com- 2 5 5 theabsenceofUVlight,whereasthephototoxicandphotonuclease plexes.Thehighoverallchargeofthecomplexesshouldfavoura agent[RhCl (dppz)(phen)]Clexhibitsnoappreciabledegreeofcell possibleinteractionwiththenegativesugar-phosphatebackbone 2 toxicitytowardsthetumorcelllinesGN4,M109andKBintheab- ofDNA. senceofirradiationat311nm[21].Uponphotoactivation,another rhodium(III) complex [Rh(bpy) 2 (chrysi)]3+ has been shown to 2.Resultsanddiscussion cleavetheDNAbackboneclosetothebindingsiteofthelargearo- maticchrysiligandatasinglebasemismatch[22].Thesefindings 2.1.Synthesisandstructuresof1–7 underlinethepossibleinfluenceoftheancillaryligandsonboththe DNAbindingandthebiologicalpropertiesofgroup9polypyridyl The compounds of the type [RhCl(pp)([9]aneS )]2+ [(pp)=bpy, 3 complexes. bpm, phen, tap, dpq, dppz] 2–7 were synthesized by heating the However,therehavebeennopreviousreportsonthebiological solventcomplexes[RhCl (pp)(L)](L=CH OH,H O)withanequiv- 3 3 2 activity and DNA interaction of cytotoxic thioether rhodium(III) alentof1,4,7-trithiacyclononaneat78(cid:3)Cinmethanol.Treatment complexes containing polypyridyl ligands. In continuation of our of RhCl (cid:2)3H O with the appropriate polypyridyl ligand (pp) in 3 2 studiesoncytotoxicrhodium(III)complexeswehavenowsynthes- either CH OH or CH OH/CH Cl afforded the complexes 3 3 2 2 ised novel complexes of the type fac-[RhCl(pp)([9]aneS 3 )]2+ 2–7 [RhCl 3 (pp)(L)]insituandtheywereemployedwithoutfurthersep- containingthetridentateS-donorligand[9]aneS 3 andthepolypyr- arationorcharacterization.Complexes2–7werecharacterizedby idylligandsbpy,bpm,phen,tap,dpq,dppz(Scheme1).Thesecom- pounds were synthesized for the purpose of evaluating their structure–activityrelationshipsandtheirinvitrocytotoxicitiesto- wardthehumancancercelllinesMCF-7,HT-29andthelymphoma celllineBJABaswellastheirinteractionwithDNA.Inanalogyto Table1 aromatic coligands such as [g5-C Me ](cid:3) or g6-C Me the ability Crystalandrefinementdataforcomplexes1and6. 5 5 6 6, of the trithia macrocycle 1,4,7-trithiacyclononane to coordinate 1 6 transitionmetalsinafacialmannerasasix-electrondonoriswell Empiricalformula C12H16Cl3N2O2Rh C20H22Cl3N4O4RhS3 documented[23,24].Ontakingstructure–activityrelationshipsof M(g/mol) 429.53 687.86 half-sandwichorganometalliccomplexesofthetype[g6-arene)Ru Temperature(K) 153 110 Cl(LL)]+(LL=en[25],pp[20])and[(g5-C Me )RhCl(pp)]+[11]into Crystalsystem Monoclinic Triclinic account,threefactorsmightbeexpected 5 top 5 ossiblyinfluencethe Spacegroup P21/c P-1 a(Å) 9.030(4) 8.0478(4) biologicalpropertiesofsuchRh(III)compounds.Theirinvitrocyto- b(Å) 14.035(7) 8.7095(5) toxicitycoulddependon(a)thelabilityoftheRh–Clbond,(b)the c(Å) 12.525(6) 19.2758(11) a((cid:3)) 90 79.852(5) b((cid:3)) 100.945(9) 85.439(5) c((cid:3)) 90 76.826(4) 2+ V(Å3) 1558.5(13) 1293.86(12) 3 Z 4 2 S 2 4 2 3 4 2 3 N F(000) 856 692 S Rh N 5 N 11 12 N N N 5 P R C c r a a y d lc s i d t a a t ( l i g o s / n c iz m e 3 ( ) mm) M 1 0. . 3 8 o 5 3 -K 1 (cid:4) a 0.30(cid:4)0.28 1 0 M . . 1 7 o 9 6 -K 6 (cid:4) a 0.18(cid:4)0.16 S Cl N 6 N 14 13 N N N 6 l 2h ( m m ax m ( (cid:3) (cid:3)) 1) 2 1 7 .6 .5 1 9 1 2 1 5 .2 .0 4 0 8 9 7 N h,k,l-ranges (cid:3)7,11/(cid:3)18,18/(cid:3)16,15 (cid:3)9,9/(cid:3)10/10/(cid:3)22,22 8 9 7 9 8 8 Collectedrefections 10503 20441 bpy 2 bpm3 tap5 Uniquereflections 3574 4561 phen 4 Observedreflections[I>2r(I)] 3060 3587 dpq 6 R1[I>2r(I)] 0.045 0.032 dppz7 wR2(alldata) 0.116 0.086 S(goodness-of-fit) 1.098 1.041 max/minDq(eÅ(cid:3)3) 1.719/(cid:3)1.50 0.93/(cid:3)0.68 Scheme1. Structuresofthecationsofthecomplexes2–7. 700 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 Fig.1. Molecularstructureofmer-[RhCl3(bpy)(CH3OH)](cid:2)CH3OH1.Selectedbond Fig. 2. Molecular structure of the cation of [RhCl(dpq)([9]aneS3)]Cl2(cid:2)4H2O 6 lengths(Å)andangles((cid:3)):Rh1–Cl12.356(1),Rh1–Cl22.322(1),Rh1–Cl32.346(1), withoutcounterionsandco-crystallizedwatermolecules.Selectedbondlengths Rh1–O11 2.103(3), Rh1–N1 2.008(3), Rh1–N2 1.989(3), Cl1-Rh1–Cl2 176.04(3), (Å)andangles((cid:3)):Rh1–S12.287(1),Rh1–S22.302(1),Rh1–S3(2.308(1),Rh1–N1 Cl3–Rh1–N1178.59(8),N2–Rh1–O11174.94(10),N1–Rh1–N281.14(12). 2.080(3),Rh1–N22.087(3),Rh1–Cl12.371(1),S1–Rh1–N2174.70(8),S2–Rh1–Cl1 176.24(3),S3–Rh1–N1175.83(8),N1–Rh1–N280.13(11). 1H NMR and positive-ion LSIMS and gave satisfactory therateofCl(cid:3)/H 2 Oexchangeofthecoordinatedchlorideligandsis microanalyses. ofprimaryinterest.TheRh1–Cl1distanceof2.371(1)Åissignifi- ReactionofRhCl (cid:2)3H Oand2,20-bipyridineinboilingmethanol cantly longer than those of 2.322(1)–2.356(1)Å for the Rh–Cl 3 2 ledtorapidformationofanorangesolidthatcontainsamixtureof bonds of 1, which is in accordance with the stronger trans influ- the fac and mer isomers of [RhCl 3 (bpy)(CH 3 OH)] with the former enceofthe[9]aneS 3 sulfuratomS2incomparisontothechloride isomerclearlypredominating.Onleavingthereactionmixtureto orbpynitrogenatomsof1.AlthoughtheRh–Nbondsof6exhibit stand the latter isomer mer-[RhCl (bpy)(CH OH)](cid:2)CH OH 1 similardistanceswithRh1–N1being2.080(3)ÅandRh1–N2being 3 3 3 (Scheme 2) slowly crystallized from the remaining solution in 2.087(3)Å, a significant difference is observed for the trans posi- the form of red prismatic crystals and its crystal structure was tioned Rh–S bonds. At 2.287(1)Å the Rh1–S1 distance is some determinedbyX-raydiffraction(Table1,Fig.1).Inthemeridional 0.02Å shorter than that of 2.308 (1)Å for the Rh1–S3 bond. The arrangementoftheligandsin1,thechloridesCl1undCl2aresited Rh1–S2distanceof2.302(1)ÅtranstotheRh1–Cl1bondissimilar in trans-position to one another. At 2.322 (1)Å the Rh1–Cl2 dis- tothatofRh1–S3.TheresultinganglesatthecentralRh1atomare tance is significantly shorter than the Rh1–Cl1 bond length of all close to 90(cid:3), with S1–Rh1–Cl1 [86.57(3)(cid:3)] and N2–Rh1–S3 2.356(1)Å, presumably as a result of the participation of Cl1 in [95.74(8)]representingthelargestdeviations. anO21–H21(cid:2)(cid:2)(cid:2)Cl1hydrogenbondoflength3.090(4)Å.Thesolvent methanoloxygenO21isalsoinvolvedinastrongerO21(cid:2)(cid:2)(cid:2)H11–O11 2.2.UV/VisandDNAbindingstudies interaction[2.648(4)Å]tothecoordinatedmethanolligand. On dissolving in d -DMSO solution, the mer isomer 1 exhibits Half-sandwich organometallic complexes of the type [(g5- 6 theexpectedsplittingofitsnonequivalentH2andH91HNMRres- C 5 Me 5 )RhCl(pp)]+(pp=dpq,dppz)arestrongside-onintercalators onancesto9.00and9.47ppm.ConfirmationofthejOcoordination into DNA [11] as are the analogous Ru(II) compounds [(g6- of the CH 3 OH ligand is provided by the downfield shift of its 1H C 6 Me 6 )RuCl(pp)]+ [20]. In contrast, covalent binding to the DNA NMR methyl resonance from 3.16ppm for the free molecule to nucleobases is thermodynamically preferred by the softer {(g5- 3.52ppm in 1. On standing in the presence of light, a slow C 5 Me 5 )Ir(pp)}2+ fragment of the complexes [(g5-C 5 Me 5 )IrCl(pp)]+ CH 3 OH/DMSO exchange is observed, leading after 18h to a following rapid Cl(cid:3)/H 2 O exchange [27]. This is also the case for 0.23:0.77 ratio of coordinated to free CH 3 OH. As previously re- (g5-C 5 Me 5 )Rh(III) and (g6-C 6 Me 6 )Ru(II) compounds containing ported, fac-[RhCl (DMSO)(bpy)] predominates in comparison to thesmallerpolypyridylligandsbpyandphen. 3 mer-[RhCl (DMSO)(bpy)]intheresultingsolution[10].Acompara- Hypochromicshiftsforcharacteristicabsorptionbandsofpoly- 3 bleCH OH/H Oexchangeandmer/facisomerisationisobservedfor pyridylmetalcomplexesandincreasesintheDNAmeltingtemper- 3 2 thecomplexinaqueoussolution.Thebulkprecipitatesoftheanal- atureT m generallyreflectthestrengthofpossibleintercalationinto ogous intermediates [RhCl (pp)(CH OH)] (pp=bpm, phen, tap, the double helix [18,19]. Following initial incubation periods of 3 3 dpq,dppz)allcontainmixturesoffacandmerisomers.Ourfind- 5min, UV/visible (UV/Vis) spectra of buffered solutions of com- ings indicate that the fac isomersare thermodynamically favour- plexes 2–7 (10mM phosphate buffer, pH=7.2) with calf thymus able in polar solvents but that the isolation of mer isomers may DNA (CT DNA) at 25(cid:3)C and r=0.1 exhibit no further significant bepossiblebyslowcrystallisationfrompolarsolutionsasdemon- changesover48h,therebyindicatingthatachievementofequilib- stratedforcomplex1.Thisisalsothecaseformer-[IrCl (DMSO)- riummustberapid.Onlyminorchangesinabsorbanceintherange 3 (phen)], which can be crystallised by slow evaporation of a of320–450nmwereobservedfor2–6onmixingwithDNA.Incon- CH OH/H Omixture,although1HNMRindicatesthattheoriginal trast,theabsorptionmaximaat362and380nmforcomplex7are 3 2 solutiononlycontainsthefacisomer[12]. shiftedby,respectively,4and2nmtohigherwavelengthandex- Acrystalstructureanalysiswasalsoperformedforthecomplex hibitDA/Avaluesof(cid:3)0.26and(cid:3)0.36thatareclearlyinaccordance [RhCl(dpq)([9]aneS )]Cl 6.Crystalsof[RhCl(dpq)([9]aneS )]Cl (cid:2)4 with dppz intercalation into DNA (Supplementary material, 3 2 3 2 H O were obtained by slow evaporation of an aqueous solution. Fig.S1). 2 Fig. 2 depicts the molecular structure (Table 1) of its cation Time-dependentUV/Visstudiesofsolutionsofmetalcomplexes [RhCl(dpq)[9]aneS )]2+ without the counter ions and the four co- inthephosphatebufferwithoutCTDNArevealedslowhypochro- 3 crystallized water molecules. Two of the four water molecules mic changes in the absorption spectra. Over a period of 48h the aredisorderedasisoneofthechloridecounterions,whosedisor- absorbance values of a solution of [RhCl(dpq)([9]aneS 3 )]Cl 2 6 at derisassociatedwiththatofoneofthewateroxygens.Forpossible 293K are significantly reduced in the range 200–270nm. At the substitutionreactionsofthemetalcomplexesinbiologicalmedia, maximum of 263nm a difference DA/A max in the absorbance of R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 701 Fig. 4. CD spectra of CT DNA and mixtures of [RhCl(bpy)([9]aneS3)]Cl2 2 and [RhCl(phen)([9]aneS3)]Cl2 4 with CT DNA {r=[complex]/[DNA]=0.1 for Fig.3. UV/Visspectraofthedppzcomplex7atthebeginningofthekineticstudyin [DNA]=M(basedonnucleotidenumber)}ina10mMphosphatebuffer(pH=7.2) a phosphate buffer (1min), and after 20h and 44h. The decrease at the three afteranincubationperiodof60min.Molarellipticities[h]aregivenintheunits maximaisshownaswellastheisosbesticpointat388nm. degcm2dmol(cid:3)1(cid:4)10(cid:3)3. (cid:3)0.11isobserved.ThisbehaviourisinaccordancewithslowCl(cid:3)/ DNAformforwhichanegativeCDbandat246nmcausedbythe H O and/or Cl(cid:3)/phosphate exchange. In contrast, effectively no helicalconformationandapositiveCDbandat275nmduetobase 2 change is determined for the maximum at 342nm. Similar rates stacking are observed [30]. Changes induced by specific interac- forthehypochromicshiftsinabsorptionareobservedfortheother tionsbetweenthebiomoleculeandthearomaticligandofthetran- complexes with smaller polypyridyl ligands. In the case of time- sition metal complexes can induce CD bands between 280 and dependent UV/Vis studies on complex [RhCl(dppz)([9]aneS )]Cl 400nmasaresultofintercalation,surfaceorgroovebinding.Dis- 3 2 7 in the phosphate buffer absorption changes were ascertained tortionsoftheinitialBconformationmayoftenbegaugedonthe forthemaximumat282aswellasat362nmand380nmovera basisofchangestothemajorbandsat246and275nm.Nosignif- periodof44h.Fig.3comparestheUV/Visspectraofthedppzcom- icant changes in the CT DNA spectra were detected for the com- plex7atthebeginningofthekineticstudyinphosphatebufferand plexes [RhCl(bpm)([9]aneS 3 )]Cl 2 3 and [RhCl(tap)([9]aneS 3 )]Cl 2 5 after20hand44h.DifferencesDA/A intheabsorbanceof(cid:3)0.25 withthesmallernitrogen-richbpmand tap ligands (Supplemen- max and(cid:3)0.27areobservedafter44hforthemaximaat362nmand tarymaterial,Fig.S2). 380nm. A smaller value of (cid:3)0.09 is found for the maximum of Toassuresolubilityofthecomplexes2,4and7,1%DMSOwas 282.5nm.Itisinterestingtonotethatanisosbesticpointisappar- addedtotheemployedaqueousbuffersolutions.TheCTDNAspec- entat388nmforthesolutionsafter20and44hbutthattheinitial tra in the presence of [RhCl(bpy)([9]aneS 3 )]Cl 2 2 and [RhCl(phe- spectrumdoesnotpassthroughthispoint.Thissuggeststhattwo n)([9]aneS 3 )]Cl 2 4 both exhibit a blue shift and a significant products may be formed, e.g. one containing an aqua ligand and decrease in molar ellipticity [h] for the positive CD band at one a phosphate ligand. To confirm the nature of the exchange 275nm,changesthatcouldindicateadistorsionoftheBDNAcon- reaction, time-dependent UV/Vis studies were also performed in formation (Fig. 4). The CD spectra of [RhCl(dppz)([9]aneS 3 )]Cl 2 7 aTris–HClbuffer(1MTrispH7.4,1.5MNaCl;pH=7.4)forcom- with CT DNA (Fig. 5) contains the expected pronounced induced plex 7 alone. In contrast to the phosphate buffer, no significant negative CD Signal at 294nm for an intercalative binding mode, changesinabsorbancewereobservedforthemaximaat362and as has been previously observed for other half-sandwich dppz 380nm, which clearly indicates that chloride substitution reac- complexes [11,20,27]. The maximum of the base stacking signal tionsmustbetakingplaceinthephosphatebuffer.Adecreasein issignificantlyshiftedtolowerwavelengthandthenegative[h]va- absorbance at the maxima at 362nm (DA/A =(cid:3)0.03) and max 380nm(DA/A =(cid:3)0.05)is,however,observedinaqueoussolu- max tionatpH=6.3.TheAvaluesremainconstantafter2h.Asnoisos- besticpointispresentinthiscaseitcanbeconcludedthatinitial fairlyrapidCl(cid:3)/H OsubstitutionisfollowedbyslowerH O/phos- 2 2 phateexchangeinthephosphatebuffer. Effectivelyunchangedorlowerthermaldenaturationtempera- tures of +1, 0, (cid:3)5, and 0(cid:3)C, respectively, wererecorded for com- plexes 2–5 and are in accordance with an absence of intercalation. In contrast, the DNA melting temperature curves for complexes 6 and 7 were discontinuous and nonreversible. Time-dependentUV/Visstudiesfor6and7aloneinthephosphate bufferat70(cid:3)CdemonstratethattheDA/A changesatthecom- max plexmaximaat258nmfor6and280nmfor7causedbytheaccel- erated rate of chloride exchange will prevent the satisfactory determinationofthehigherT valuesofDNAinthepresencesof m thesecomplexes. Inthepresenceofsmallmolecules,characteristicchangesinthe Fig. 5. CD spectra of CT DNA and mixtures of [RhCl(dpq)([9]aneS3)]Cl2 6 and circulardicroism(CD)spectraofDNAintherangeof220–400nm [RhCl(dppz)([9]aneS3)]Cl2 7 with CT DNA {r=[complex]/[DNA]=0.1 for [DNA]=M(basedonnucleotidenumber)}ina10mMphosphatebuffer(pH=7.2) can provide a means of monitoring possible conformational afteranincubationperiodof60min.Molarellipticities[h]aregivenintheunits changes for the biopolymer [28,29]. CT DNA is present in the B degcm2dmol(cid:3)1(cid:4)10(cid:3)3. 702 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 lueoftheinducedcirculardicroism(ICD)signalat245nmforthe helix conformation is reduced by about 40%. These significant changes in the CD absorption of the biopolymer are clearly in accordance with intercalation of the dppz ligand. In the case of complex 6 with the smaller polypyridyl ligand dpq even more drasticchangescanbedetected(Fig.5). The detected exiton with two ICD signals at 280 and 297nm indicates the presence of two competing modes of DNA binding [29].Groovebindingorintercalationareprobablebindingmodes for6withDNAbutcovalentbindingfollowingachlorideexchange mayalsobepossible.Thesignificantblueshiftofthebasestacking signalandthepresenceofnegativeICDsignalsindicate,inanalogy Fig. 7. Gel retardation assay of plasmid pBR322 (20lM in base pairs) in the to the dppz complex 7, at least partial intercalation. Significant presenceof2lMconcentrationsofcomplexes2and6afterirradiationfor1hat changesinthehelixconformationmustalsoinvolvedasindicated 311,342and366nm. bythe17%reductioninthenegative[h]valueoftheCDsignalat 245nm. Head-on intercalation from the minor groove has been suggestedfor(dpq)Ru(II)complexesonthebasisofNOESYexper- comparison toethidiumbromide.For complex2withits smaller iments [31], but the side-on mode is more typical for half-sand- bpm ligand, the observed effects are less pronounced. At lower wichpolypyridylcomplexes[18,19]. concentrationsof2lMand5lM(Fig.6b,lines5and6)muchof the supercoiled DNA is converted into nicked DNA. In line 7 for 2.3.Nucleaseactivity the higher complex concentration of 20lM, the intensity of the detectionagentethidiumbromideisweakerandthebottomline Light induced nuclease activity can offer a potent means of alsoexhibitsalowerretardationthanthesupercoiledlineofplas- selectively treating cancer through local irradiation and thereby midpBR322alone,asinthecaseofcomplex6at5lMconcentra- activation of an applied agent. Photochemical cleavage of DNA tion.AnindicationofmultipleDNAstrandbreaksisalsogivenby has been reported for the complexes [(g5-C Me )Ir(L)(dppz)]2+ theweakerintensityofthisline. 5 5 (L=Ac-Met-OMe, H-Gly-Gly-Met-OH, H-Met-OH=L-methionine, Toascertaintheoptimalconditionsforactivationofthephoto- H-Gly-OH=glycine) [18] and rhodium intercalators [21,32]. We, nucleasepropertiesof2and6,irradiationexperimentswereper- therefore,analyzedpossibleDNAcleavagecausedbyselectedcom- formedforvariouswavelengthsandirradiationtimes.Irradiation plexesthroughelectrophoresisinanagarosegel.Incubationinthe of a mixture of complex 6 (2lM) with plasmid pBR322 (20lM) dark of plasmid pBR322 (Fermentas) with the complexes atawavelengthof311nmoverperiodsof0.5,1.0and1.5hleads [RhCl(bpy)([9]aneS )]Cl 2 and [RhCl(dpq)([9]aneS )]Cl 6 under inallcasestocompleteconversionofsupercoiledDNAintonicked 3 2 3 2 aerobic conditions led only in the case of latter dpq complex (at DNA(resultsnotshown).Threewavelengths(311nm,342nmand concentrationsof2–20lM)toconversionofsupercoiledDNAinto 366nm) were selected on the basis of the absorption maxima of nickedDNA,albeitatarelativelymoderatelevel(Fig.6a,lines2–4). the complexes in their UV spectra. Short wavelength UV light In the case of bpy complex 2, no significant differences between (311–325nm) is often used to produce initial strand breaks near the plasmid alone and the mixtures of plasmid with complex 2 the metal complex binding site [15]. The 311nm wavelength in wereobservedatsimilarconcentrations(Fig.6a,lines5–7). this range is located close to an absorption shoulder of complex Incubation of the complexes with the plasmid pBR322 under 6 at 303nm and to an absorption maximum of complex 2 at roomlightirradiationfor24hled,incontrast,tosignificantdiffer- 312nm. The wavelength 342nm represents a minor absorption ences. The presence of light has no effect on the plasmid alone maximum (MLCT) for complex 6 and low absorbance values are (Fig.6b,line1),butataconcentrationofonly2lMcomplex6con- observed for both complexes at 366nm. After irradiation for 1h vertsmuchofthesupercoiledDNAintonickedDNA(Fig.6b,line at the various wavelengths (311nm, 342nm and 366nm) little 2).Atahigherconcentrationof5lM(Fig.6b,line3)nearlyallof cleavagewasdetectedfortheplasmidalone(Fig. 7,lines3–5).A thesupercoiledDNAisconvertedintonickedDNAandinaddition GeneRulerTM1kb DNAladder(Fermentas) isshownin lines 1and thebottomlineexhibitsalowerretardationthanthesupercoiled 13as a reference.Forcomplex 2 a numberofimportantchanges lineofplasmidpBR322alone.Thisindicatesthepresenceofalin- can be recognized. Whereas on irradiation at 366nm very little earorshorterfragmentthanforsupercoiledDNA.Theintensityof supercoiledDNAisconvertedintonickedDNA(line8),irradiation thenickedfragmentismuchweakerthanforthelowerconcentra- at342nm(line7)causesabouthalfthesupercoiledDNAtobecon- tionofcomplex6inline2.Atahighercomplexconcentrationof verted.Atthelowerwavelengthof311nm(line6)almostallofthe 20lMmultipleDNAstrandbreaksareprovokedandnosignificant supercoiledDNAisconvertedtonickedDNAtogetherwithalittle fragmentsaredetectablebygelelectrophoresis(Fig.6b,line4).In linearDNAwithamolecularmassofabout5000bp.Irradiationof the case of photoinduced intercalation, the intensity signal could complex6withplasmidpBR322(line9)at311nmleadstocom- alsobe decreased as aresultofstrongerbindingof complex6in plete cleavage of supercoiled DNA and converts this into nicked and linear DNA with molecular masses of about 6500 and 5000bp. The nicked DNA in the presence of complex 6 is about 500bpshorterthanthenormalvalueofabout7000bpfornicked DNA (lines 3–5) on its own. On irradiation at 342nm only the nickedformisobservedatabout6500bp,whereasafterirradiation at 366nm both supercoiled and nicked DNA (6500bp) are ob- servedataboutthesameintensity.TheinducedDNAstrandscis- sion caused by irradiation of the plasmid in the presence of Fig. 6. Gel retardation assays of plasmid pBR322 DNA (20lM in b.p.) after compound6isnotaffectedbyanexcessofDMSO(ahydroxylrad- incubationfor24hwithcomplexes2and6:(a)inthedarkand(b)withroomlight icalscavenger)andonlytoaminorextentbysodiumazide(asin- irradiation.Inbothcasesline1containsplasmidpBR322,lines2–4plasmidpBR322 pluscomplex6atconcentrationsof2lM,5lMand20lMandlines5–7plasmid gletoxygenquencher)andCatalasebovineserum,whichsuggests pBR322pluscomplex2atconcentrationsof2lM,5lMand20lM,respectively. anoxygen-independentcleavagemechanism[32,33]. R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 703 The photocleavage assays confirm the induction of nuclease the compounds 2, 4, 6 and 7, which contain polypyridyl ligands properties for the ([9]aneS )Rh(III) complexes by UV light. ofdifferingsizes,anincreaseoftheantiproliferativepotencywith 3 Compound2canberegardedasthemoreeffectivephotonuclease increasing surface area and hydrophobicity of the polypyridyl li- becauseofitslackofcleavagepropertiesinthedark.Theexperi- gandswastobeexpected[20].Whereastheactivitiesof4,6and ments also confirm good cleavage results for wavelengths, that 7do,indeed,followtheexpectedtrend,thebpycomplex2surpris- arelocatedneartotheabsorptionmaximaofthecompounds. inglyexhibitscellgrowthinhibitingpropertiescomparabletothat of the dppz species 7. The high cytotoxic activity of 2.4.Reactivitytowardssmallbiomolecules [RhCl(dppz)([9]aneS )]Cl 7 withIC values of 4.7 (±0.5) and 7.4 3 2 50 (±2.2)lM towards MCF-7 and HT-29 cells is comparable to that TheobservationoftwonegativeICDsignalsintheCDspectrum ofcisplatin[2.0(±0.3)and7.0(±2.09)lM][20]. ofCTDNAin thepresenceofcomplex[RhCl(dpq)([9]aneS )]Cl 6 Promisingresultswerealsoobtainedwiththecomplexes3and 3 2 suggests that a second binding mode may compete with partial 5,whichcontainbpmandtapasligands.AstheNdonoratomsof intercalationofthedpqligand.We,therefore,studiedthepossible ligandstapandbpmwillexhibitincreasedlocalisedelectronden- reactionofguanosine50-monophosphatewith6(2:1)inthedarkat sityincomparisontobpyandphentheyshouldbemoreeffective pH=6.15by1Hand31PNMRspectroscopy.However,afteranincu- inreducingthepositivechargeonthecentralrhodiumatomand bationperiodof24hat37(cid:3)CnoadditionalsignalsforN7-coordi- therebylabilizingtherhodium–chloridebond.Asignificantdepen- natedproducts wereobserved. This lackof reaction isin striking denceoftheinvitrocytotoxicityonthenumberofaromaticnitro- contrast to [(g5-C Me )RhCl(dpq)](CF SO ) for which formation genatomsisapparentforthesmallerpolypyridylligands.TheIC 5 5 3 3 50 ofthejN7productisrapid[11].Thepresenceofanew31PNMR values for [RhCl(bpm)([9]aneS )]Cl 3 are comparable to those of 3 2 signal at 10.5ppm in comparison to that at 2.4ppm for free cisplatin and are the lowest of the series with respectively 1.7 50-GMP2(cid:3)is,however,indicativeofadegreeofphosphatecoordi- (±0.5)and1.9(±0.1)lMforMCF-7andHT-29cells.Furthermore, nation (14%). We also detected no significant reaction for a 1:2 [RhCl(tap)([9]aneS )]Cl 5 exhibits with IC values of 11.5 (±0.6) 3 2 50 solution of complex [RhCl(dpq)([9]aneS )]Cl 6 with N-acetylme- and7.5(±4.8)lMforMCF-7andHT-29cellsamuchhigheractivity 3 2 thionine (Ac-Met-OH) over the same period of time. Substitution thanthatoftheanalogousphencomplex4.Theincreasedcytotoxic ofthechlorideligandbythethioethersulfuratomofAc-Met-OH activityof3and5couldpossiblyresultfromanincreaseintherate mightbeexpectedtoleadtoadownfieldshiftforthe1HNMRsig- ofchlorideexchange.Ontheotherhand,thepresenceofadditional nal of the adjacent d-CH group from 2.18 to about 2.40pm, but nitrogen atoms may also favour molecular recognition mecha- 3 this was not observed. However an ESI MS peak of low relative nismsowingtothepossibilityoftheseatomsparticipatingin X– abundance (10%) was detected for the molecular ion [{Rh(Ac- H(cid:2)(cid:2)(cid:2)N (X=N,O) hydrogen bonds. In summary, the bpy complex 2 Met-O)(dpq)([9]aneS )}-H]+atm/z703.8uintheaqueousreaction displaysahigheractivitythanthecomplexes4and6withthelar- 3 mixture6/Ac-Met-OHafter30daysat25(cid:3)Cinadditiontothatof gerphenanddpqligandsandthecomplexes3and5ofthenitro- [{RhCl(dpq)([9]aneS )}-H]+ at 549.0 u (100%). This suggests that gen-rich polypyridyl ligands bpm and tap are significantly more 3 complex6maypossiblyundergoalimiteddegreeofCl(cid:3)/Ac-Met- cytotoxicthantheir2,20-bipyridineand1,10-phenanthrolinebased O(cid:3)exchangeandthatthe{([9]aneS )Rh(dpq)}3+fragmentmaybe counterparts. 3 preferentially coordinated by the harder carboxylate oxygen of Itisinterestingtonote,thatthecoligand[9]aneS aloneshows 3 Ac-Met-OH rather than its softer thioether sulfur atom, a finding nocytotoxiceffecttowardsthehumancancercelllinesMCF-7and thatwouldbeinaccordancewiththeknownborderlineproperties HT-29 (IC values >100lM). The IC values [11] for bpy alone 50 50 ofrhodium(III)atoms[34]. [52.7 (±7.8) and 45.7 (±4.6)lM] are much larger towards MCF-7 Takentogetherwiththeobservationofsignificantchlorideex- andHT-29cellsthanforcomplex2.Incontrastfree1,10-phenan- change in a phosphate buffer at pH=7.2, these reactivitystudies throlineexhibitsasignificantlyhighercytotoxicity[3.5(±0.2)and suggest thatthe complexes2–7 maypreferentially undergo slow 2.7 (±0.5)lM] than the phen compound 4 towards these human substitutionreactionswithhard Odonor atomsratherthan with cancercelllines.Theinfluenceofthemetalfragmentisparticularly the softer N or S atoms of biomolecules such as oligonucleotides strikingforthenitrogen-richtapligand.Whereastapaloneexhib- orpeptides. its no significant cytotoxicity [IC value>100lM for MCF-7], its 50 complex5is3.2timesmoreactivetowardsMCF-7and9.6times 2.5.Cytotoxicitymeasurementsandcellularuptake moreactivetowardsHT-29cellsthanthephencomplex4. Cellular uptake studies (Table 2) were performed exemplarily In vitro cytotoxic studies of compounds 2–7 were carried out for the selected compounds 2, 3 and 5. The accumulation of the withthehumanMCF-7 (breast cancer)and HT-29(coloncancer) complexesinthetumorcellsislowincomparisontootherstruc- cell lines and the resulting IC values are listed in Table 2. For turallyrelatedrhodiumcompounds[10,11]andacorrelationwith 50 Table2 IC50(lM)andcellularuptake(ngRh/mgprotein)valuesforthecomplexes[RhCl(pp)([9]aneS3)]2+2–7. Compound pp Counterions MCF-7IC50 HT-29IC50 MCF-7uptake10lM HT-29uptake10lM 2 bpy Cl 12.8(0.2) 4.4(0.1) 4.62(1.21) 2.08(0.18) 3 bpm Cl 1.7(0.5) 1.9(0.1) 2.79(1.41) 3.13(0.20) 4 phen PF6 36.3(6.0) 72.2(8.0) n.d. n.d. 5 tap Cl 11.5(0.6) 7.6(4.8) 3.70(0.04) 1.34(0.28) 6 dpq Cl 19.1(0.3) 20.9(2.8) n.d. n.d. 7 dppz Cl 4.7(0.5) 7.4(2.2) n.d. n.d. [9]aneS3 >100 >100 bpy[11] 52.7(7.8) 45.7(4.6) phen[11] 3.5(0.2) 2.7(0.5) tap >100 n.d. cisplatin[20] 2.0(0.3) 7.0(2.0) n.d.:Notdetermined. 704 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 the IC values could not be established. It is interesting to note 2–60lM and 7 at concentrations in the range 9–25lM (Supple- 50 thatparticularlycomplex3displayedverypromisingcytotoxicef- mentarymaterial,Fig.S3).Theseresultsindicatethatnecrosisdoes fects despite its low uptake into the tumor cells. Thus, a drug nothaveasignificantimpactontheactivityofthecomplexes. designaimingatnovelstructurallyrelatedcomplexeswithanim- BJABcellsincubatedwithincreasingconcentrationsoftherho- proveduptakeprofilemightleadtocompoundswithenhancedcell diumcompoundsexhibitedreducedproliferationandaloweredcell growthinhibitingpotency. viabilityafteranincubationperiodof24hconfirmingtheabovede- scribedantiproliferativepotentialofthecomplexes.Apoptosis,in 2.6.Apoptosisinduction contrasttounspecificnecrosis,requiresacontrolledandregulated mechanismleadingtocelldeath.DNAfragmentation(hypoploidy [RhCl(bpm)([9]aneS )]Cl 3 and [RhCl(dppz)([9]aneS )]Cl 7 isconsideredtobeanadditionalindicatorofapoptoticcelldeath 3 2 3 2 wereselectedasthemostactivecompoundstoascertainwhether and we, therefore, quantified the induction of apoptosis via flow theobservedcelldeathtriggeredbytherhodiumcompoundswas cytometric measurements of the DNA fragments after incubating duetoapoptosisornecrosis.Necroticcelldeathischaracterizedby BJABcellsfor72hwith3and7[35].ThenumbersofapoptoticBJAB theearlyreleaseoflactatedehydrogenase(LDH),whereasapopto- cellsfordifferentconcentrationsof3areillustratedinFig.8.Asmay ticcells,incontrast,retaintheirmembraneintegrityanddonotex- beseen,significantDNAfragmentationisobservedevenatlowcon- hibitanearlyreleaseoflargeintracellularproteinssuchasLDH.No centrationsofthecytotoxiccomplex. significant LDH release was observed for lymphoma cells (BJAB) Our investigations also clearly demonstrate that complexes 3 after1hincubationperiodswith3atconcentrationsintherange and 7 trigger the mitochondrial pathway of apoptosis. As illus- trated in Fig. 9, dose dependent loss of the mitochondrial mem- brane potential was observed for BJAB cells after 48h of incubationwiththerhodium(III)compounds.Complex7isclearly more efficient in this respect. At a concentration of 20lM the dppz compound 7 is some 25% more effective than the bpm compound 3. After staining the cells with a dye JC-1 (5,50,6,60- tetrachloro-1,1,3,30-tetraethyl-benzimidazolylcarbocyanineiodide), themitochondrialpermeabilitywasquantifiedbyflowcytometric determination of the cells with decreased fluorescence, i.e. those containingmitochondriadisplayingalowermembranepotential. 3.Conclusions The observation of significantly higher cytotoxicity towards MCF-7 and HT-29 cells for both [RhCl(bpy)([9]aneS )]2+ 2 and 3 Fig.8. DNAfragmentationforBJABcellsafterincubationfor72hwithdifferent [RhCl(dppz)([9]aneS )]2+ 7 in comparison to [RhCl(phe- 3 concentrations of 3. Data are given in % hypoploidy (subG1) which reflects the n)([9]aneS )]2+ 4 suggests that differing mechanisms of action numberofapoptoticcells. 3 and target molecules may be involved for the ([9]aneS )Rh(III) 3 complexes, depending on whether these contain either small or large polypyridyl ligands. Whereas no specific interaction with DNAcouldbeestablishedfor2oritsmoreactive2,20-bipyrimidine analogue 3 both 6 (pp=dpq) and 7 exhibit strong intercalative binding.Thismodeofinteractionleadstomoderatenucleaseactiv- ityfor6inthedark,whereasUVirradiationisrequiredtoinduce DNAcleavageinthepresenceof2.Thenatureofcelldeathforlym- phoma cells (BJAB) was studied for the most active complexes 3 and7.Bothinducesignificantapoptosisviatheintrinsicmitochon- drialpathway.Theconsiderablepotentialofthese([9]aneS )Rh(III) 3 polypyridylcomplexesascytotoxicagentsisfurtherunderlinedby thefactthatboth3and7causeonlynegligiblenecroticcelldeath inBJABcells. 4.Experimental 4.1.Materialsandmethods UV/visible(UV/Vis)spectrawererecordedwithanAnalytikJena SPECORD 200 spectrometer and CD spectra with a Jasco J-715 instrumentintherange220–400nmfor1:10complex/[DNA]mix- tures [complex=20lM, DNA concentration in M(nucleo- tide)=200lM]ina10mMphosphatebufferatpH7.2.Toassure the solubility of complexes 2, 4 and 7, 1% DMSO was added to the aqueous buffer. LSIMS spectra (LSIMS=liquid secondary ion Fig. 9. Mitochondrial permeability transition as measured by flow cytometric mass spectrometry) were registered for the mass range m/ analysisforBJABcellsafter48hincubationwithdifferentconcentrationsof3and7. z=3000 with a Fisons VG Autospec employing a cesium ion gun Valuesofthemitochondrialpermeabilitytransitionaregivenaspercentagesofcells (voltage 17kV) and 3-nitrobenzyl alcohol as the liquid matrix. A withlowDW m±esd(n=3).JC-1=5,50,6,60-tetrachloro-1,10,3,30-tetraethyl-benzim- idazolylcarbocyanineiodide. Bruker DRX 400 and a Bruker DPX 200 were employed for the R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 705 registrationof1HNMRspectrawithchemicalshiftsreportedasd sequentlywashedanddriedinvacuo.Yield:64.7mg(28%).Anal. values relative to the signal of the deuterated solvent. The split- Calcd.forC H Cl N O Rh S (601.83):N9.31,C27.94,H4.02,S 14 24 3 4 3 1 3 tingsofprotonresonancesinthereported1HNMRspectraarede- 15.98%.Found:N9.4,C27.2,H4.0,S16.5%.Althoughtheanalysisva- fined as s=singlet, d=doublet, t=triplet and m=multiplet. lueforCwaslessthansatisfactory,additionalphysicaldata,espe- Elemental analyses were performed on a Vario EL of Elementar ciallytheLSIMSand1HNMRresultsareclearlyconsistentwiththe Analysensysteme GmBH. RhCl (cid:2)3H O was obtained from ABCR, chemicalformulaforcompound3.LSIMS:m/z(%)=511(8)[M(cid:3)Cl]+, 3 2 2,20-bipyrimidinefromAlfaAesar,1,10-phenanthroline,2,20-bipyr- 475(100)[M(cid:3)Cl(cid:3)HCl]+,441(17.8)[M(cid:3)Cl(cid:3)2HCl]+.1HNMR(D O, 2 idine, pyrazino[2,3-f]quinoxaline (tap), Ethidium bromide, NaN 400MHz, 25(cid:3)C): d=3.33–3.97 (m, 12H, CH –[9]aneS ), 8.14 (t, 3 2 3 from Acros Organics, and 1,4,7-trithiacyclononane, calf thymus 4J=10.8Hz,2H,bpm-H3/H8),9.38(d,3J=5.9Hz,2H,bpm-H4/H7), DNA (CT DNA), Catalase bovine serum from Sigma Aldrich. The 9.42(d,3J=4.9Hz,2H,bpm-H2/H9). plasmidpBR322andthegeneruler1kbDNAladderwereobtained fromFermentas.Theligandsdpq[31]anddppz[36]wereprepared 4.2.4.[RhCl(phen)([9]aneS )][PF ] (cid:2)2H O(4) 3 62 2 inaccordancewithliteratureprocedures. 1,10-phenanthroline(68.5mg,0.38mmol)wasaddedtoasolu- tionofRhCl (cid:2)3H O(100.0mg,0.38mmol)in10mL1:1methanol/ 3 2 4.2.Synthesisofcomplexes1–7 dichloromethane and the reaction mixture was refluxed for 2h. Following addition of 68.4mg (0.38mmol) 1,4,7-trithiacyclonon- 4.2.1.mer-[RhCl (bpy)(MeOH)](cid:2)CH OH(1) ane and refluxing for 10h a yellow precipitate of 3 3 RhCl (cid:2)3H O (100.0mg, 0.38mmol) was dissolved in 10mL ([RhCl ([9]aneS )]) was removed. After solvent removal from the 3 2 3 3 methanol and 2,20-bipyridine (59.3mg, 0.38mmol) was added remainingsolution, the residue was redissolved in methanol and and refluxed for 3h. After a few min an orange solid precipitate additionofanexcessofNH PF ledtoprecipitationoftheproduct, 4 6 ofpredominantlyfac-[RhCl (bpy)(MeOH)]wasformed.Onleaving whichwaswashedanddriedinvacuo.Yield:142mg(47%).Anal. 3 the suspension to stand for 12h red crystals were also obtained Calcd. for C H Cl F N O P Rh S (824.87): N 3.4, C 26.21, H 18 24 1 12 2 2 2 1 3 from the solution. After manual separation from the remaining 2.93, S 11.66%. Found: N 3.7, C 25.6, H 2.9, S 11.5%. LSIMS: m/z powder these were characterized by X-ray structural analysis as (%)=642 (84) [M(cid:3)PF ]+, 496.9 (100) [M-PF6-HPF6]+, 461.4 (24) 6 mer-[RhCl (bpy)(MeOH)](cid:2)CH OH 1: Yield: 15.2mg (9.4%). Anal. [M-2HPF (cid:3)Cl]+.1HNMR(D O,400MHz,25(cid:3)C):d=3.36–3.96(m, 3 3 6 2 Calcd. for C H Cl N ORh(cid:2)CH OH (428.52): N 6.54, C 33.63, H 12H,CH [9]aneS ),8.23(q,2H,H3/8phen),8.37(s,2H,H12/13 11 11 3 2 3 2 3 3.53%. Found: N 6.5, C 32.9, H 3.8%. 1H-NMR (DMSO-d , phen),9.03(q,2H,H4/7phen),9.35(d,2H,H2/9phen). 6 200MHz, 25(cid:3)C): d=3.53 (s, 3H, MeOH ), 7.88 (m, 2H, bpy-H3/ Rh H8), 8.30 (m, 2H, bpy-H4/H7), 8.69 (t, 2H, bpy-H5/H6), 8.93 (d, 4.2.5.[RhCl(tap)([9]aneS )]Cl (cid:2)2.5H O(5) 3 2 2 1H, bpy-H2), 9.61 (d, 1H, bpy-H9); 1H-NMR (D O, 200MHz, Preparationof5wasdoneasfor3exceptthattheligandpyraz- 2 25(cid:3)C): d=3.67 (s, 3H, MeOH ), 7.84 (m, 2H, bpy-H3/H8), 8.29 ino[2,3–f]quinoxaline(tap)(68.5mg,0.38mmol)wasused.Yield: Rh (m, 2H, bpy-H4/H7), 8.48 (m, 2H, bpy-H5/H6), 9.00 (d, 1H, bpy- 49.3mg (21%). Anal. Calcd. for C H Cl N O Rh S (616.84): N 16 23 3 4 2.5 1 3 H2),9.47(d,1H,bpy-H9). 9.08, C 31.15, H 3.76, S 15.6%. Found: N 8.9, C 31.1, H 3.8, S 15.7%. LSIMS: m/z (%)=536.2 (1.5) [M(cid:3)Cl]+, 499 (100) 4.2.2.[RhCl(bpy)([9]aneS )]Cl (2) [M(cid:3)Cl(cid:3)HCl]+, 465 (30) [M(cid:3)Cl(cid:3)2HCl]+. 1H NMR (D O, 400MHz, 3 2 2 2,20-bipyridine(59.3mg,0.38mmol)wasaddedtoasolutionof 25(cid:3)C): d=3.4–4.0 (m, 12H, CH , [9]aneS ), 8.76 (s, 2H, tap-H5/ 2 3 RhCl (cid:2)3H O (100.0mg, 0.38mmol) in 20mL methanol/dichloro- H6),9.53(d,2H,tap-H3/H8),9.6(d,2H,tap-H2/H9). 3 2 methane(1:1).Afterrefluxingthemixturefor2hinboilingmeth- anol/dichloromethane, 1,4,7-trithiacyclononane (68.4mg, 4.2.6.[RhCl(dpq)([9]aneS )]Cl (cid:2)3H O(6) 3 2. 2 0.38mmol)wasaddedandthesuspensionwasstirredforafurther Dipyrido[3,2-d:20,30-f]quinoxaline (88.1mg (0.38mmol) was 2hinboilingCH OH/CH Cl .AfterevaporationofCH Cl ,thereac- addedtoRhCl (cid:2)3H O(100.0mg,0.38mmol)in40mL1:1CH Cl / 3 2 2 2 2 3 2 2 2 tionmixturewasstirredforafurther2hinboilingmethanol.Fol- MeOHandthemixturewasstirredfor3hinboilingCH OH/CH Cl . 3 2 2 lowing removal of a yellow precipitate of [RhCl ([9]aneS )] by Afteradditionof[9]aneS (68.4mg,0.38mmol),thesuspensionwas 3 3 3 filtrationfromthereactionmixture,theremainingsolutionwasre- refluxedforafurther19h.Followingfiltrationoftheyellowprecip- ducedinvolumeto5mL.Theorangeproductwasprecipitatedby itate ([RhCl ([9]aneS )]) and solvent removal from the remaining 3 3 addition of diethyl ether and subsequently washed with CH OH red solution, the resulting residue was redissolved in methanol. 3 and dried in vacuo. Yield: 61.4mg (30%). Anal. Calcd. for Additionofdiethyletherledtoprecipitationoftheproduct,which C H Cl N RhS (545.81): N 5.13, C 35.21, H 3.69, S 17.62%. was washed with CH OH and dried in vacuo. Yield: 169.2mg 16 20 3 2 3 3 Found: N 5.0, C 35.7, H 4.4, S 17.8%. LSIMS: m/z (%)=510.0 (6.7) (72%). Anal. Calcd. for C H Cl N O Rh S (675.91): N 8.29, C 20 26 3 4 3 1 3 [M(cid:3)Cl]+, 475.0 (13.4) [M(cid:3)Cl(cid:3)HCl]+, 329.1 (100) 35.54,H3.88,S14.23%.Found:N8.3,C35.6,H3.9,S13.6%.LSIMS: [M(cid:3)Cl(cid:3)[9]aneS ]+. 1H NMR (D O, 200MHz, 25(cid:3)C): d=3.34–3.87 m/z (%)=587 (8.2) [M(cid:3)Cl]+, 549 (100) [M(cid:3)Cl(cid:3)HCl] +, 515 (22) 3 2 (m, 12H, [9]aneS ) 7.93 (t, 4J=3.0, 2H, bpy-H3/H8), 8.44 (t, [M(cid:3)Cl(cid:3)2HCl]+. 1H NMR (D O, 200MHz, 25(cid:3)C): d=3.39–4.0 (m, 3 2 4J=3.4, 2 H, bpy-H4/H7), 8.68 (d, 3J=4.2, 2 H, bpy-H5/H6), 8.99 12H,CH ,[9]aneS ),8.39(q,2H,H3/8dpq),9.29(s,2H,H12/13, 2 3 (d,3J=2.9,2H,bpy-H2/H9). dpq),9.46(d,2H,H4/7,dpq),9.89(d,2H,H2/9,dpq).Singlecrystals of[RhCl(dpq)([9]aneS )]Cl (cid:2)4H OsuitableforX-rayanalysiswere 3 2 2 4.2.3.[RhCl(bpm)([9]aneS )]Cl (cid:2)3H O(3) grownbyslowevaporationofanaqueoussolution. 3 2 2 2,20-bipyrimidine (60.1mg, 0.38mmol) dissolved in 10mL methanol was added to a solution of RhCl (cid:2)3H O (100.0mg, 4.2.7.[RhCl(dppz)([9]aneS )]Cl (cid:2)(C H ) O7 3 2 3 2 2 5 2 0.38mmol) in 25mL methanol and the mixture was refluxed for Preparation as for 6 with one equivalent of the ligand dipyr- 2h.Theinitialredsolutionyieldedaredsuspensionwithinthefirst ido[2,3-a:20,30-c]phenazine (107.3mg, 0.38mmol). Dichlorometh- 10min of reaction. 68.4mg (0.38mmol) 1,4,7-trithiacyclononane ane was allowed to evaporate before addition of the ligand weresubsequentlyaddedandthesuspensionwasstirredforafur- [9]aneS .Yield:79.7mg(34%).Anal.Calcd.forC H Cl N ORh S 3 28 32 3 4 1 3 ther2hinboilingmethanol.Afterremovalofayellowprecipitate (746.04):N7.51,C45.08,H4.32,S12.89%.Found:N7.2,C45.1,H of [RhCl ([9]aneS )] by filtration from the reaction mixture, the 4.7, S 13.0%. LSIMS: m/z (%)=599 (100) [M(cid:3)Cl(cid:3)HCl]+, 565 (17.6) 3 3 remainingsolutionwasreducedinvolumeto5mL.Theyellow-or- [M(cid:3)Cl(cid:3)2HCl]+. 1H NMR (D O, 200MHz, 25(cid:3)C): d=3.32–3.63(m, 2 angeproductwasprecipitatedbyadditionofdiethyletherandsub- 12 H, CH [9]aneS ), 8.06–8.1 (m, 2H, H 13/14 dppz), 8.29–8.32 2 3 706 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 (m,4H,H3/8,H12/15dppz),9.43(d,3J=5.8,2H,H4/7dppz),9.7– for96h(MCF-7)or72h(HT-29)thecellbiomasswasdetermined 9.8(m,2H;H2/9dppz). bycrystalvioletstainingand theIC valuesweredeterminedas 50 thoseconcentrationscausing50%inhibitionofcellproliferation.Re- 4.3.X-raystructuralanalysisof1and6 sultswerecalculatedfrom2to3independentexperiments. Cytotoxicityof3and7towardsBJABcellswasmeasuredbyre- Crystalandrefinementdatafor1and6aresummarisedinTable leaseoflactatedehydrogenase(LDH)asdescribedpreviously[40]. 1.Intensitydatawerecollectedusing1(cid:3)xscansandMo-Karadi- Afterincubationwithdifferentconcentrationsofthecomplexesfor ation (k=0.71073Å) on a Bruker SMART diffractometer at 153K 1hor3hat37(cid:3)C,LDHreleasedbyBJABcellswasmeasuredinthe for 1 and with an Oxford Diffraction Xcalibur2 diffractometer cellculturesupernatantsusingtheCytotoxicityDetectionKitfrom equippedwithaSapphire2CCDat110Kfor6.Thedatafor1and Boehringer Mannheim(cid:4) (Mannheim, Germany). The supernatants 6werecorrectedempirically(SADABS)for absorptionand solved were centrifuged at 1500rpm for 5min. Cell-free supernatants bydirectmethodswithSHELX97.RefinementagainstF2 wasper- (20lL) were diluted with phosphate-buffered saline (PBS, 80lL) 0 formed by SHELXL97 [37] with anisotropic temperature factors and a reaction mixture containing 2-[4-iodophenyl]-3-[4-nitro- for non-hydrogen atoms and protons at geometrically calculated phenyl]-5-phenyltetrazolium chloride (INT), sodium lactate, positions as riding atoms. The final R factors for 1 were NAD+anddiaphorase(100lL)wasadded.Time-dependentforma- R =0.045 for 3060 reflections with I>2r (I) and wR = tion of the reaction product was quantified photometrically at 1 2 [Rw(F2(cid:3)F2)2/Rw(F 2)2]1/2=0.116, S (goodness-of-fit)=1.098 for 490nm. The maximumamountof LDH released by the cells was 0 c 0 all3574independentreflections.Theanalogousvaluesfor6were determinedbylysisofthecellsbyusing0.1%TritonX-100incul- R =0.032 for 3587 reflections (I>2r (I)), wR =0.086 and turemediumandsetas100%celldeath. 1 2 S=1.041forall4561independentreflections. 4.7.DeterminationofcellconcentrationandcellviabilityforBJABcells 4.4.DNAbindingstudiesof2–7 Cell viability of the BJAB cells was determined by using the TheUV/Viskineticstudiesandthermaldenaturationtempera- CASY(cid:4) Cell Counter+Analyzer System from Innovatis (Bielefeld, tureT determinationsfor1:10complex/DNAmixtures[DNAcon- Germany).Settingswerespecificallydefinedfortherequirements m centration=M(basedonnucleotidenumber)]wereperformedina of the used cells. With this system, the cell concentration can be 10mMphosphatebufferatpH=7.2.Meltingcurveswererecorded analyzedsimultaneouslyinthreedifferentsizeranges:celldebris, at2(cid:3)Cstepsforthewavelength260nmwithanAnalytikJenaSPE- dead cells, and viable cells. Cells were seeded at a density of CORD200spectrometerequippedwithaPeltiertemperaturecon- 1(cid:4)105cellsml(cid:3)1 and treated with different concentrations of 3 troller.DT valueswerecalculatedbydeterminingthemidpoints and7;non-treatedcellsservedascontrols.Aftera24hincubation m ofmeltingcurvesfromthefirst-orderderivatives.Theexperimen- period at 37(cid:3)C, cells were resuspended properly and 100lL of talDT valuesareestimatedtobeaccuratewithin±1(cid:3)C.Concen- eachwellwasdilutedin10mLCASYton(ready-to-useisotonicsal- m trationsofCTDNAweredeterminedspectrophotometricallyusing inesolution)foranimmediateautomatedcountofthecells. themolarextinctioncoefficiente =6600M(cid:3)1cm–1[38]. 260 4.8.Cellularuptakestudies 4.5.Cellcultures Forcellularuptakestudies,MCF-7andHT-29cellsweregrown MCF-7 breast cancer and HT-29 human colon carcinoma cells until at least 70% confluency in 175cm2 cell culture flasks [41]. weremaintainedin10%(v/v)fetalcalfserumcontainingcellculture Stock solutions of complexes 2–7 in DMSO or H O were freshly 2 medium (minimum essential medium eagle supplemented with preparedanddilutedwithcellculturemediumtothedesiredcon- 2.2gNaHCO ,110mg/Lsodiumpyruvateand50mg/Lgentamicin centrations (final DMSO concentrations: 0.1% v/v, final complex 3 sulfateadjustedtopH 7.4)at37(cid:3)C/5% CO andpassagedtwicea concentration:10lM).Thecellculturemediumofthecellculture 2 week according to standard procedures. BJAB (burkitt-like lym- flasks was replaced with 10mL of the cell culture medium solu- phoma)weremaintainedat37(cid:3)CinRPMI1640(GIBCO,Invitrogen) tions containing 2–7 and the flasks were incubated for 6h at supplemented with 10% heat inactivated fetal calf serum, 37(cid:3)C/5% CO . Afterwards the culture medium was removed, the 2 100,000UL(cid:3)1 penicillin, 0.1gL(cid:3)1 streptomycin, and 0.56gL(cid:3)1 L- cell layer washed with 10mL PBS (phosphate buffered saline pH glutamine.Thecellsweresubculturedevery3–4daysbydilution 7.4), treated with 2–3mL trypsin solution (0.05% trypsin, 0.02% of the cells to a concentration of 1(cid:4)105cellsmL(cid:3)1. Twenty-four EDTA in PBS) and incubated for 2min at 37(cid:3)C/5% CO after re- 2 hoursbeforetheassaysetup,cellswereculturedataconcentration moval of the trypsin solution. Cells were resuspended in 10mL of3(cid:4)105cellsmL(cid:3)1toascertainstandardizedgrowthconditions. PBSandcellpelletswereisolatedbycentrifugation(roomtemper- Forapoptosisassays,thecellswerethendilutedtoaconcentration atures,2000g,5min).Theisolatedcellpelletswereresuspendedin of1(cid:4)105cellsmL(cid:3)1immediatelybeforeadditionofthedifferent 1–5mLtwicedistilledwater.Therhodiumcontentofthesamples complexes. wasdeterminedbygraphitefurnaceatomabsorptionspectrome- try (GF-AAS) (Section 4.9) and the protein contents of separate 4.6.Cytotoxicitymeasurements aliquots by the Bradford method. To correct for matrix effects in GF-AAS measurements, samples and standards were adjusted to Theantiproliferativeeffectsofthecompoundsweredetermined the same protein concentration by dilution with twice-distilled following an established procedure [39]. In short, cells were sus- water.PriortoGF-AASanalysis,20lLtritonX-100(1%)and20lL pended in cell culture medium (MCF-7: 10,000cells/mL, HT-29: nitricacid(13%)wereaddedtoeach200lLsampleofthecellsus- 2850cells/mL),and100lLaliquotsthereofwereplatedin96well pension.Cellularuptakewasexpressedasngrhodiumpermgcell platesandincubatedat37(cid:3)C/5%CO for72h(MCF-7)or48h(HT- proteinfordataobtainedfromthreeindependentexperiments. 2 29).StocksolutionsofthecompoundsinDMSOorH Owerefreshly 2 preparedanddilutedwithcellculturemediumtothedesiredcon- 4.9.Atomicabsorptionspectrometry centrations(finalDMSOconcentration:0.1%v/v).Themediumin theplateswas replacedwithmediumcontainingthecompounds AVariographitefurnaceatomicabsorptionspectrometer(Anal- in gradedconcentrations(sixreplicates).After furtherincubation ytik Jena) was employed for the Rh quantification using a R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 707 wavelengthof343.5nm withabandpassof 0.5nm.A deuterium 1.5mL Eppendorf tubes using a 150W Xenon lamp plus mono- lampwasemployedforthebackgroundcorrection.Matrixcontain- chromator (Oriel). Samples (10lL) were irradiated for 1h and ingstandardswereobtainedbyadditionofrhodiumstocksolution stored at (cid:3)20(cid:3)C. Controls without metal complexes were also (1mgmL(cid:3)1 Rh in 5% HCl) to untreated cell suspensions. Probes testedinparallelforlightdamage.DNAsingleanddoublestrand were injected at a volume of 20lL into standard graphite tubes. breaksweredetectedbygelelectrophoresisassaysusing1%aga- Drying,pyrolysisandatomizationwereperformedaccordingtolit- rosegelswith10mMTAE(Tris–acetate–EDTA)astherunningbuf- erature conditions [11]. During the temperature programme the fer(pH=8.4).Gelswereruninthedarkfor45–75minat100Vand graphite tube was purged with a constant argon gas flow, which stained with aqueous ethidium bromide (0.5lg/ml). They were wasonlyhaltedduringthezeroingandatomisationsteps.Pyroly- documented by a Gel documentation station PEQLab BioVersion sis and atomisation temperatures were optimised prior to the 3000andquantifiedwithBio1Dsoftware. experiments. The recovery rates of the metallodrugs were deter- mined initially and used for calculation of the final experimental 5.Abbreviations values.Themeanintegratedabsorbancesofduplicatedinjections were used thoughout the study. The characteristic concentration forthedescribedmethodwas0.85±0.05(lgRhL(cid:3)1)/1%A. [9]aneS 1,4,7-trithiacyclononane 3 bpy 2,20-bipyridine 4.10.MeasurementofDNAfragmentationofBJABcells bpm 2,20-bipyrimidine tap pyrazino[2,3-f]quinoxaline Apoptoticcelldeathwasdeterminedbya modifiedcellcycle- phen 1,10-phenanthroline analysis,whichdetectsDNAfragmentationatthesinglecelllevel dpq dipyrido[3,2-d:20,30-f]quinoxaline asdescribedpreviously[35,42].Cellswereseededatadensityof dppz dipyrido[2,3-a:20,30-c]phenazine 1(cid:4)105cellsmL(cid:3)1 and treated with different concentrations of 3 dppn benzo[i]dipyrido[3,2-a:20,30-c]phenazine and7.Aftera72hincubationperiodat37(cid:3)C,cellswerecollected en ethylenediamine [9]aneNS 1-aza-4,7-dithiacyclononane bycentrifugationat1500rpmfor5min,washedwithPBSat4(cid:3)C 2 [12]aneN 1,4,7,10-tetraazacyclododecane andfixedinPBS/2%(v/v)formaldehydeonicefor30min.Afterfix- 4 [12]aneS 1,4,7,10-tetrathiacyclododecane ation,cellswerepelleted,incubatedwithethanol/PBS(2:1,v/v)for 4 phi 9,10-diaminophenanthroline 15min, pelleted and resuspended in PBS containing 40lgmL(cid:3)1 tpy 2,20:6,200-terpyridine RNaseA(Qiagen,Hilden,Germany).RNAwasdigestedfor30min (RR)-Me trien 2R,9R-diamino-4,7-diazadecane 37(cid:3)C, after which the cells were pelleted once again and finally 2 MLCT metal-ligandchargetransfer resuspendedinPBScontaining50lgmL(cid:3)1propidiumiodidefrom LSIMS Liquidsecondaryionmassspectrometry Serva (Heidelberg, Germany). Nuclear DNA fragmentation was LDH lactatedehydrogenase quantified by flow cytometric determination of hypodiploid DNA INT 2-[4-iodophenyl]-3-[4-nitrophenyl]-5-phenyltetrazolium (Fluorescence–activatedcellsort,FACS).Datawerecollectedand chloride analyzedusingaFACScan(BectonDickinson,Heidelberg,Germany) JC-1 5,50,6,60-tetrachloro-1,10,3,30-tetraethyl-benzimidazolyl- apparatusequippedwiththeCELLQuestsoftware.Dataaregivenin carbocyanineiodide %hypoploidy(subG1),whichreflectsthenumberofapoptoticcells PBS phosphatebufferedsaline GF-AAS graphite-furnaceatomabsorptionspectrometry 4.11.Measurementofthemitochondrialpermeabilitytransition TAE tris–acetate–EDTA TE tris–EDTA Afteranincubationperiodof48hwithdifferentconcentrations of3and7,thecellswerecollectedbycentrifugationat1500rpm, 4(cid:3)Cfor5min.Themitochondrialpermeabilitytransitionwasthen Acknowledgements determinedbystainingthecellswith5,50,6,60-tetrachloro-1,10,3,30- tetraethyl-benzimidazolylcarbocyanine iodide (JC-1; Molecular Financial support of this work in Bochum and Berlin by the Probes, Leiden,The Netherlands). 1(cid:4)105 cells wereresuspended DeutscheForschungsgemeinschaft(DFG)withintheresearchgroup in 500lL phenol red-free RPMI 1640 without supplements and FOR 630 ‘‘Biological Function of Organometallic Compounds” is JC-1 was added to give a final concentration of 2.5lgmL(cid:3)1. The gratefully acknowledged. Ruth Bieda wishes to thank both the cells were incubated for 30min at 37(cid:3)C with moderate shaking. Ruhr-University Research School and the WASAG Foundation for Control cells were likewise incubated in the absence of JC-1 dye. theawardofscholarships.WearealsogratefultoHeikeScheffler, The cells were harvested by centrifugation at 1500rpm, 4(cid:3)C for ManuelaWinterandUlrikeSchmidtforexcellenttechnicalsupport. 5min, washed with ice-cold PBS and resuspended in 200lL PBS at4(cid:3)C.Mitochondrialpermeabilitytransitionwasthenquantified AppendixA.Supplementarydata byflowcytometricdeterminationofthecellswithdecreasedfluo- rescence,i.e.withmitochondriadisplayingalowermembranepo- CCDC-701717and(cid:3)701718containthesupplementarycrystal- tential.DatawerecollectedandanalyzedusingaFACScan(Becton lographicdatafor1and6andmaybeobtainedfreeofchargeat Dickinson, Heidelberg, Germany) apparatus equipped with the http://www.ccdc.cam.ac.uk/conts/retrieving.html or from the CELLQuestsoftware.Dataaregivenin%cellswithlowDW m ,which Cambridge Crystallographic Data Centre, 12 Union Road, Cam- reflectsthenumberofcellsundergoingmitochondrialapoptosis. bridge CB2 1EZ, UK; fax: +44 1223/336x033; e-mail: deposit@ccdc.cam.ac.uk. 4.12.Gelelectrophoresis Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,atdoi:10.1016/j.jinorgbio.2009.01.008. Samples were prepared in 10mM TE buffer (10mM Tris–HCl, 1mM EDTA, pH 7.6) and contained 20lM (bp) pBR322 double- References strandedcircularplasmidDNA(c=4361lmoll(cid:3)1inbp)andvari- ousconcentrationsof2lM,5lMor20lMoftherhodiumcom- [1] M.A. Jakupec, M. Galanski, V.B. Arion, C.G. Hartinger, B.K. Keppler, Dalton plexes 2 and 6. Irradiation were carried out successively in Trans.(2008)183–194. 708 R.Biedaetal./JournalofInorganicBiochemistry103(2009)698–708 [2] L.Ronconi,P.J.Sadler,Coord.Chem.Rev.251(2007)1633–1648. [24] K.Brandt,W.S.Sheldrick,J.Chem.Soc.DaltonTrans.(1996)1237–1243. [3] P.J.Dyson,G.Sava,DatonTrans.(2006)1929–1933. [25] Y.W.Yan,M.Melchart,A.Habtemariam,P.J.Sadler,Chem.Commun.(2005) [4] N.Katsaros,A.Anagnostopoulou,Crit.Rev.Oncol.Hematol.42(2002)297– 4764–4776. 308. [26] S.C.Rowle,G.Admans,S.R.Cooper,J.Chem.Soc.DaltonTrans.(1988)93–96. [5] A.M. Angeles-Boza, H.T. Chifotides, J.D. Aquirre, A. Chouai, P.K.L. Fu, K.R. [27] S.Schäfer,W.S.Sheldrick,J.Organomet.Chem.692(2007)1300–1309. Dunbar,C.Turro,J.Med.Chem.49(2006)6841–6847. [28] V.Brabec,V.Kleinwächter,J.-L.Butour,N.P.Johnson,Biophys.Chem.25(1990) [6] P.Clu,H.He,X.Liu,Progr.Chem.18(2006)1646–1651. 129–141. [7] F.P.Pruchnik,P.Jakimowicz,Z.Ciunik,J.Zakrzewska-Czerwinska,A.Opolski,J. [29] M.Eriksson,B.Norden,MethodsEnzymol.340(2001)68–98. Wietrzyk,E.Wojdat,Inorg.Chim.Acta334(2002)59–66. [30] W.C. Johnson, in: N. Berova, K. Nakanishi, R.W. Woody (Eds.), Circular [8] D.A.Medvetz,K.D.Stakleff,T.Schreiber,P.D.Custer,K.Hindi,M.J.Panzner,D.D. Dichroism: Principles and Applications, second ed., VCH Publishers, New Blanco,M.J.Taschner,C.A.Tessier,W.J.Youngs,J.Med.Chem.50(2007)1703– York,2000,pp.523–540. 1706. [31] J.G. Collins, A.D. Sleemann, J.R. Aldrich-Wright, I. Greguric, T.W. Hambley, [9] B. Serli, E. Zangrando, T. Gianferrara, C. Scolaro, P.J. Dyson, A. Bergamo, E. Inorg.Chem.37(1998)3133–3141. Alessio,Eur.J.Inorg.Chem.(2005)3423–3434. [32] C.G. Barry, E.C. Turney, C.S. Day, G. Saluta, G.L. Kucera, U. Bierbach, Inorg. [10] M.Harlos,I.Ott,R.Gust,H.Alborzinia,S.Wölfl,A.Kromm,W.S.Sheldrick,J. Chem.41(2002)7159–7169. Med.Chem.51(2008)3924–3933. [33] B.Armitage,Chem.Rev.98(1998)1171–1200. [11] M.A.Scharwitz,I.Ott,Y.Geldmacher,R.Gust,W.S.Sheldrick,J.Organomet. [34] D.A.Herebian,C.S.Schmidt,W.S.Sheldrick,C.vanWüllen,Eur.J.Inorg.Chem. Chem.693(2008)2299–2309. (1998)1991–1998. [12] M.A.Scharwitz,I.Ott,R.Gust,A.Kromm,W.S.Sheldrick,J.Inorg.Biochem.102 [35] T.Wieder,A.Prokop,B.Bagci,F.Essmann,D.Bernicke,K.Schulze-Osthoff,B. (2008)1623–1630. Dörken, H.-G. Schmalz, P.T. Daniel, G. Henze, Leukemia 15 (2001) 1735– [13] K.E.Erkkila,D.T.Odom,J.K.Barton,Chem.Rev.99(1999)2777–2795. 1742. [14] C.Metcalfe,J.A.Thomas,Chem.Soc.Rev.32(2003)215–224. [36] A.Delgadillo,P.Romo,A.M.Leiva,B.Loeb,Helv.Chim.Acta86(2003)2110– [15] B.M.Zeglis,V.C.Pierre,J.K.Barton,Chem.Commun.(2007)4565–4579. 2120. [16] C.L.Kielkopf,K.E.Erkkila,B.P.Hudson,J.K.Barton,D.C.Rees,Nat.Struct.Biol.7 [37] G.M.Sheldrick,SHELXS97andSHELXL97,Göttingen,Germany,1997. (2000)117–121. [38] H.-Q.Liu,T.-C.Cheung,S.-M.Peng,C.-M.Che,J.Chem.Soc.Chem.Commun. [17] T.P.Shields,J.K.Barton,Biochemistry34(1995)15049–15056. (1995)1787–1788. [18] D.Herebian,W.S.Sheldrick,J.Chem.Soc.DaltonTrans.(2002)966–974. [39] I.Ott,K.Schmidt,B.Kircher,P.Schumacher,T.Wiglenda,R.Gust,J.Med.Chem. [19] A.Frodl,D.Herebian,W.S.Sheldrick,J.Chem.Soc.DaltonTrans.(2002)3664– 48(2005)622–629. 3673. [40] R.A.Diller,H.M.Riepl,O.Rose,C.Frias,G.Henze,A.Prokop,Chem.Biodiversity [20] S.Schäfer,I.Ott,R.Gust,W.S.Sheldrick,Eur.J.Inorg.Chem.(2007)3034–3046. 2(2005)1331–1337. [21] E.L. Menon, R. Perera, M. Navarro, R.J. Kuhn, H. Morrison, Inorg. Chem. 43 [41] I.Ott,H.Scheffler,R.Gust,Chem.Med.Chem.2(2007)702–707. (2004)5373–5381. [42] P.K. Smith, R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H. Gartner, M.D. [22] B.A.Jackson,V.Y.Alekseyev,J.K.Barton,Biochemistry38(1999)4655–4662. Provenzano, E.K. Fujimoto, N.M. Goeke, B.J. Olson, D.C. Klenk, Analyt. [23] S.R.Cooper,Acc.Chem.Res.21(1988)141–146. Biochem.150(1985)76–85.