Cytotoxicity, hydrophobicity, uptake, and distribution of osmium(II) anticancer complexes in ovarian cancer cells.

840 J.Med.Chem.2010,53,840–849 DOI: 10.1021/jm901556u Cytotoxicity, Hydrophobicity, Uptake, and Distribution of Osmium(II) Anticancer Complexes in Ovarian Cancer Cells SabineH.vanRijt,ArindamMukherjee,AnaM.Pizarro,andPeterJ.Sadler* DepartmentofChemistry,UniversityofWarwick,GibbetHillRoad,CoventryCV47AL,U.K. ReceivedOctober20,2009 Thecytotoxicity,hydrophobicity(logP),cellularuptake,aqueousreactivity,andextentofDNAadduct formation in the A2780 ovarian carcinoma cells for four osmium(II) arene complexes [(η6-arene)Os- (4-methyl-picolinate)Cl]thatdifferonlyintheirareneligandsasbenzene(1),p-cymene(2),biphenyl(3), ortetrahydroanthracene(4)arereported.Thereisacorrelationbetweenhydrophobicity(logP),cellular uptake,nucleusuptake,andcytotoxicityofthecomplexes,followingtheorder3∼4>2>1,suggesting that the arene plays an important role in the biological activity of these types of compounds. Cell distributionstudiesusingfractionationshowedthatallfourcompoundsdistributesimilarlywithincells. DNA binding of osmium did not correlate with cytotoxicity, indicating that the nature of the DNA lesionmayalsobecrucialtoactivity.TEMimagesofovariancellstreatedwith3revealedmorphological changesassociatedwithapoptosiswithpossibleinvolvementofmitochondria. Introduction oforganometallicOsIIarenecomplexesappeartohavepro- misinginvitrocancercellactivityandtheiraqueousreactivity Anticanceragentsusuallyexerttheirtherapeuticeffectby appearstobetunable.17-22Inaddition,someoftheseosmium their interaction with intracellular targets. However, the complexesarenon-cross-resistantwithcisplatintowardcan- limitedpenetrationofcytotoxicdrugsintotumorsisbelieved cercells,20,22suggestingthattheyhaveadifferentmechanism to be a major contributing factor to the frequent failure of of action and show potential for addressing the problem of chemotherapy tocompletelyeradicate tumorsinthe clinic.1 intrinsicoracquiredresistanceinchemotherapy. For example, in mice, the anticancer drug doxorubicin Here we have studied four osmium(II) arene complexes only diffuses 40-100 μm from blood vessels and shows limited penetration into the solid tumors.2 The metal-based [(η6-arene)Os(4-methyl-picolinate)Cl] that differ only in theirη6-coordinatedareneligand:benzene(1),p-cymene(2), chemotherapeutic cisplatin is taken up into cells via active transportinvolvingthecoppertransporterCTR1.3,4Unfor- biphenyl (3), and tetrahydroanthracene (4) (Figure 1). We havesoughttocorrelatetheirdifferencesinaqueousreactivity tunately, cisplatin is also exported out of the cell via active and hydrophobicity with their cytotoxicity, cellular uptake, transportinvolvingmultidrug-resistantproteins(MRP2and DNAbinding,andosmiumdistributionincellfractions. In possiblyMRP3),aswellasthecoppertransporters(ATP7B and possibly ATP7A).5,6 Reduced uptake resulting from addition, a TEMa imaging study on complex 3 allowed either increased efflux or reduced influx contributes toward observationofthedistributionofosmiuminovariancancer acquiredresistanceoftumorswithcisplatin.7-9Aneffective cellsandprovidedinsightintopotentialintracellulartargets. anticancerdrugmustreachalloftheviablecellsinatumor This report is the first detailed study of cell uptake of anti- andberetainedinsufficientconcentrationsandonarelevant cancer OsII arene complexes and reveals the importance of time scale to inhibit its intracellular target and initiate cell mitochondriaaspossibletargets. death. Anticancer drugs based on other transition metals may Results addresstheproblemsassociatedwiththeplatinumcytotoxics Synthesis and Characterization. Osmium arene com- currently used in the clinic. A number of metal-based com- pounds 1-4 were prepared via their respective Cl-bridged pounds with promising antiproliferative activity toward a dimers, [(η6-arene)OsCl ] , where arene = benzene (bz), 22 widerangeoftumorswithnovelmechanismsofactionhave p-cymene(p-cym),biphenyl(bip),andtetrahydroanthracene recentlybeendescribed.10-16Theanticancerpotentialofthe (THA), using previously reported methods.23,24 The com- third-row transition metal osmium has only recently been poundswerepurifiedbycrystallizationfrommethanoland explored. Osmium complexes have a reputation for being characterizedby1HNMRspectroscopy,ESI-MS,andele- eithertoxic(OsO )orsubstitution-inert(manyOsIIandOsIII 4 mentalanalysis(seeSupportingInformationfordetails). complexes),perhapsexplainingwhytheirtherapeuticpoten- tialhasbeenlittleinvestigated.Encouragingly,certainclasses aAbbreviations:TEM,transmissionelectronmicroscopy;ESI-MS, electrosprayionizationmassspectrometry;IC ,half-maximuminhibi- 50 *Towhomcorrespondenceshouldbeaddressed.Phone:(þ44)024 toryconcentration;ICP-MS,inductivelycoupledplasmamassspectro- 76523818.Fax:(þ44)02476523819.E-mail:p.j.sadler@warwick.ac.uk. metry. pubs.acs.org/jmc PublishedonWeb12/15/2009 r2009AmericanChemicalSociety Article JournalofMedicinalChemistry,2010,Vol.53,No.2 841 Figure2. ThelogPvaluesforcomplexes1-4.Resultsarethemean ofsixindependentexperimentsandareexpressedasmean(SD. Table2. HydrolysisDataforCompounds1-4 compd k0,h-1a t ,h 1/2 1 1.63(0.14 0.42(0.04 Figure1. Chemical structures of the organometallic osmium(II) 2 1.80(0.16 0.39(0.04 arenecomplexesstudiedinthiswork. 3 0.71(0.01b 0.98(0.02b 4 1.36(0.17 0.51(0.07 ak0isthepseudo-first-orderrateconstantdeterminedforhydrolysis Table1. InVitroGrowthInhibitionofA2780CellsforCompounds of0.8mM1-4indeionizedwaterat288Kby1HNMR.bFromref20. 1-4andCisplatin(CDDP)asControl compd IC (μM)a bondsareoftenmorereactivethanOs-Clbonds.Thismay 50 be the rate-limiting step in reactions with targets such as 1 32.7(0.8 2 7.6(0.3 DNA.25 3 3.2b Toinvestigatetheaqueousreactivityofthesecomplexes, 4 4.5(0.1 the rates of hydrolysis of compounds 1, 2, and 4 in a 5% CDDP 2.2(0.1 MeOD-d /95%D Oweremonitoredby1HNMRat288K 4 2 aDrug-treatmentperiodwas24h.Eachvaluerepresentsthemean( by the observation of new peaks over time due to aqua SDforthreeindependentexperiments.bFromref20. adductformation.Thehalf-lifeofcompound3hasrecently been reported to be t =0.98 ( 0.02 h under the same 1/2 InVitroGrowthInhibition.Allfourcomplexes1,2,3,and4 conditions.20Thepercentageofaquapeakformationfor1, (Figure 1) exhibit activity in the A2780 ovarian cancer cell 2,and4wasplottedagainsttimeandwasfittedtopseudo- linewithIC valuesof3-33μM(Table1).TheIC value first-order kinetics (Figure S1), and their half-lives were 50 50 forcomplex3isfromourrecentreport(IC =3.2μM,under calculated (Table 2). Compounds 1, 2, and 4 hydrolyzed 50 thesameconditions).20Potencyfollowstheorder4∼3> with half-lives ranging from 0.39 to 0.51 h. The extent of 2.1.Theleastactiveisthebenzenecomplex.Theextended hydrolysis at equilibrium was 100% for 1, 80% for 2, and arene derivatives have IC 50 values of 3-7 μM, compar- 60%for4(FigureS1)andwaspreviouslyreportedtobe96% able to that of drug carboplatin using a similar protocol for3.20Thehydrolysisratedeterminedforcompound1isa (IC 50 =6μM). true pseudo-first-order rate since complete hydrolysis was Hydrophobicity (logP). To assess a possible correlation observed,whilefortheothercompoundstheraterepresents between the biological activity and hydrophobicity of the thepseudo-first-orderratefortheapproachtoequilibrium osmium arene complexes, the partition coefficients (logP) sincehydrolysiswasincomplete. in an octanol-water system were determined (Figure 2). Theeffectsofchlorideconcentrationstypicalofthoseof Doublydistilledwatercontaining300mMsodiumchloride bloodplasma(100mM),cellcytoplasm(22.7mM),andcell wasusedinordertosuppresshydrolysisofthecompounds, nucleus(4mM)ontheaqueousequilibriumofcompounds1, ensuring the determination of the logP of the chlorido 2, and 4 were investigated. 1H NMR spectra of the com- complexes. The logP values increase in the order 1<2< pounds (1 mM) were recorded after enoughtime had been 4<3with-0.51(0.12for1,0.22(0.04for2,0.71(0.22 left for equilibration (at least four times their respective for 4, and 0.86 ( 0.02 for 3. These logP values are as half-lives). These same experiments have previously been expected; hydrophobicity increases with increasing size of reportedforcompound3.20 theareneligand.However,thetwo-ringsystem,bip,showed Compound 1 was present 100% as its aqua species at 4 aslightlyhigherlogPvaluethanthethree-ringsystem,with mMand22.7mMNaClconcentrations;at100mMNaCl, THA as the arene ligand. Only complex 1, containing the 47%waspresentastheaquaadduct.Forcompounds2,3, benzeneareneligand,hasanegativelogPvalue(partitions and4,atthelowNaClconcentrationof4mM,56%(2),81% preferentiallyintowater,Figure2). (3), and 67% (4) were hydrolyzed, at 22.7 mM NaCl, AqueousReactivity.Onepossiblerouteforinvivoactiva- 44% (2), 53% (3), and 50% (4) were hydrolyzed, and at tion of complexes of the type [(η6-arene)M(XY)Cl], where 100mMNaCl,2%(2),20%(3),and0%(4)werehydrolyzed M is OsII or RuII, and XY is a chelating ligand, in- (Table 3); i.e., hydrolysis is likely to be almost completely volvesaquation(replacementofClbyH O),sinceOs-OH suppressedforcomplexes2,3,and4inextracellularmedia. 2 2 842 JournalofMedicinalChemistry,2010,Vol.53,No.2 vanRijtetal. Table 3. Percentage of Aqua Adduct Formation in an Equilibrated Solution of 1 mM 1-4 in DO at Chloride Levels Typical of Blood 2 Plasma(100mM),CellCytoplasm(22.7mM),andCellNucleus(4mM) %aquaadduct compd 4mMNaCl 22.7mMNaCl 100mMNaCl 1 100 100 47 2 56 44 2 3 81a 53a 20a 4 67 50 0 aFromref20. Table4. CellularandDNAOsmiumConcentrationsinA2780Ovarian Cellsa cellularuptake DNAbinding Figure4. Cellular osmium concentrations determined in A2780 (pmolOs/106cells) (pmolOs/106cells) cellsafterexposureto5μMcomplex3for24hfollowedby24h Oscomplex mean SD mean SD in drug-free medium. Results are the mean of three independent experimentsandareexpressedasmean(SD. 1 26.2 3.3 0.9 0.2 2 59.2 13.4 0.6 0.3 intracellularlevelsofosmiumwereobservedafterexposure 3 94.1 8.2 1.3 0.3 tocomplexes4(99(9pmolOs/106cells)and3(94(8pmol 4 99.1 9.3 4.9 0.9 Os/106cells),followedbycomplexes2and1,withlevelsof aDrug-treatmentperiodwas24hwith5μMOsarenecomplexes. 59 ( 13 pmol Os/106 cells and 26 ( 3 pmol Os/106 cells, Eachvaluerepresentsthemean(SDforsixindependentexperiments. respectively. Forcomplex3,thecellularuptakeafter24hofexposure followedbya24hrecoveryperiodindrug-freemediumwas determined(Figure4).Attheendofthe24hdrugexposure, thecellularuptakewasconsistentwithpreviousexperiments. Onedayafterthedrugremoval,theamountofosmiumtaken up by the cells had decreased by about 55% (from 90 ( 7 pmolOs/106cellsto40(4pmolOs/106cells,Figure4). The cellular uptake in A2780 cells for complex 3 as a function of time (t=1, 2, 4, 8, 12, 24, and 48 h) was investigated (Figure 5). The maximum cell uptake was reached after 12 h of exposure (97 ( 5 pmol Os/106 cells). After24and48hofexposureto3,theamountofosmium decreasedto86(12pmolOs/106cellsand71(10pmolOs/ 106 cells, respectively (Figure 5A). For the first 8 h of exposureto3, thecell countwasstillincreasing. However, after12and24hofexposure,thecellcountdecreasedand after48habout80%fewercellswerecountedcomparedto thosepresentafter12hofexposure(Figure5B). DNA Adduct Formation. As for chemotherapeutic cis- platin, DNA is believed to be the main target for this type of osmiumarenecomplexes.26Forthisreason,DNA from A2780cellswasisolatedandthelevelsofosmiumonDNA were determined. The concentrations of osmium on the DNAfollowatrenddifferentfromthatobservedforcellular uptake.Compound4showsthehighestlevelofosmiumon DNA(4.9(0.9pmolOs/106cells)followedbycompounds 3,1,and2(1.3(0.3,0.9(0.2,and0.6(0.3pmolOs/106 cells, respectively) (Figure 3B). These values correspond 2 F 4 ig h ur o e f 3 e . xp O o s s m ur i e um to c 5 o μ n M cen c t o r m at p io le n x s es de 1 t - er 4 m in in ( e A d ) in wh A o 2 le 78 c 0 el c ls el a l n s d af ( t B er ) to1.0-4.9%ofosmium-DNAadductformationcompared isolatedDNA.Resultsarethemeanofsixindependentsamplesand to the total amount of osmium that was taken up by the areexpressedasmean(SD. wholecell(i.e.,4.9%for4,1.4%for3,3.4%for1,and1.0% for2). UptakeintoA2780Cells.Theuptakeofcompounds1-4 Distribution of Osmium in Cell Fractions. The osmium bytheA2780ovariancancercelllinewasstudiedtoinvesti- contentinnuclei,cytosol,andmembranefractionsisolated gate a possible relationship between the cellular uptake, from A2780 cells after 24 h of exposure to complexes 1-4 hydrophobicity,andinvitrocytotoxicityofthecomplexes. was determined, and the results are shown in Table 5 and BothcellularandDNAosmiumconcentrationsweredeter- Figure 6.The totalamounts of osmiumin thenuclear and minedafter24hofexposuretothecomplexesat5μM,which cytosoliccellfractionsfollowthesametrendaswasobserved isanaverageoftheIC valuesofcomplexes2-4.Theresults for whole-cell uptake; the total amount of osmium in the 50 are summarized in Table 4 and Figure 3A. The highest fractionsdecreasesintheorder3∼4>2>1.Forallfour Article JournalofMedicinalChemistry,2010,Vol.53,No.2 843 Figure7. TEM images of A2780 cells exposed to complex 3, showingthedifferentstagesofcellapoptosis:(A)cellcontraction; (B)membraneblebbing;(C)DNAfragmentation;(D)theforma- tionofapoptoticbodies. complexes, the amount of osmium detected accounted for about 58-66% of the total osmium taken up by the cells (Table5).Thiscanbeattributedtothelossofosmiumdur- ing the fraction-separation procedures. The distribution of Figure5. (A) Cellular osmium concentrations determined in osmiumobservedfortheOsIIcompoundsinthenuclearand A2780 cells after 1, 2, 4, 8, 12, 24, and 48 h exposure to 5 μM cytosolicfractionsisverysimilarforallfourcomplexes;with complex3and(B)cellcountsforthedifferenttimepoints.Results 14%(1),11%(2),17%(3)and14%(4)ofthetotalamount are the mean of three independent samples and are expressed as ofosmiuminthefractionslocalizinginthecellnucleusand mean(SD. with 76% (1), 72% (2), 64% (3), and 79% (4) of osmium foundinthecytosolicfractions. Table 5. Nucleus, Cytosol, and Membrane/Cytoskeleton Osmium Osmiumwasalsofoundinthecellmembrane/cytoskele- ConcentrationsinA2780OvarianCellsa tonfractions,withespeciallyhighvaluesforcomplexes2and uptake(pmolOs/106cells) 3of7.5(0.9pmolOs/106cells(17%)for2,and11.3(0.9 membraneand pmol Os/106 cells (19%) for 3, respectively (Table 5, cytosol nucleus cytoskeletonretention Figure6). Oscomplex mean SD mean SD mean SD DistributionofOsmiumUsingTransmissionElectronMi- croscopy (TEM). The high electron density of osmium is 1 12.6 2.3 2.3 0.4 1.8 0.1 2 31.1 6.6 4.7 0.9 7.5 0.9 frequently exploited for staining biological samples in var- 3 38.6 3.8 10.5 0.9 11.3 0.7 iousformsofelectronmicroscopy.Inparticular,OsO isa 4 4 46.9 5.1 8.4 1.0 4.4 0.9 widelyusedstainingagentinTEMandprovidescontrastto aDrug-treatmentperiodwas24hwith5μMOsarenecomplexes. theimage.Itseemedlikelythatwewouldobserveosmium- Eachvaluerepresentsthemean(SDforsixindependentexperiments. derived contrast insections of cancer cells treated with the complexes studied here if their deposition is sufficiently localizedincellorganelles. A2780cellswereexposedto5or20μMbiphenylcomplex 3for12h,andthetreatedandcontrolcellswerefixedand embedded in Epon resin. Ultrathin sections were observed under the TEM (Figures 7-10 and Figures S2-S4). The controlcellsandthetreatedcellswerestainedonlywith2% uranyl acetate. Any additional contrast observed between thecontrolcellsandthetreatedcellsisduetocelluptakeof complex3. Inthecellsexposedtocomplex3,morphologicalchanges associated with cell apoptosis were identified: cell contrac- tion (Figure 7A, Figure S2A,B), cell membrane blebbing (Figure7B,FigureS2C,D),DNAfragmentation(Figure7C, Figure S2E,F), and the formation of apoptotic bodies Figure6. Osmium concentration in nucleus, cytosol, and mem- branefractions(pmolOs/106cells)inA2780ovariancellsafter24h (Figure7D,FigureS3). ofexposureto5μM1-4.Resultsarethemeanofsixindependent Inthecontrolcells,thenucleusandmitochondriacanbe experimentsandareexpressedasmean(SD. observedwithrelativelylittlecontrast.Thecontrolcellsdo 844 JournalofMedicinalChemistry,2010,Vol.53,No.2 vanRijtetal. Figure8. (A-D) TEM images of A2780 untreated control cells where(B)and(D)aremagnificationsof(A)and(B),respectively. Figure10. (A-F) TEM images of A2780 cells after 12 h of ex- posureto20μMcomplex3.Osmiumcomplex3accumulatesmainly in the mitochondria, nucleolus, and nuclear membrane. Also, swelling and different morphology of the mitochondria are ob- served (D,inset,E,andF).Theinsets in(A),(C), and(D)show magnifications. appearstoaccumulateinthemitochondria,andswellingof the mitochondria seems to occur. In addition, some mito- chondria show different morphology with circular cristae insteadofthelamella-likestructuresobservedinsomeofthe othermitochondria(Figure10D-FandFigureS6D-F). Discussion Figure9. (A-D)TEMimagesofA2780cellsafter12hofexposure to5μMcomplex3.Osmiumcomplex3accumulatesmainlyinthe In this work we have investigated possible relationships mitochondria,nucleolus,andnuclearmembrane.Theinsetsin(A) betweenthehydrophobicity,cellularuptake,andcytotoxicity and(C)showmagnifications. intheA2780ovariancancercelllineoffourorganometallic osmium(II) complexes [(η6-arene)Os(4-methyl-picolinate)Cl] notappeartobedistressed,andthecellmembranesareintact which have different arene ligands, ranging from a single (Figure8andFigureS4). benzene ring (1), p-cymene (2), to the two phenyl rings of Sectionsofcellstreatedwith5μMcomplex3showmore biphenyl (3) and three fused rings of tetrahydroanthracene contrastincomparisontothosefromcontrolcells,inparti- (4).Inaddition,time-dependentcelluptake,thedistributionof cularinthemitochondria,nucleolus,andthenuclearmem- osmiumincellfractions,DNAbinding,andTEMimagingof brane (Figure 9, Figure S5). Several cells displayed the cell sections were studied to gain insight into the mechanism morphologicalchangesassociatedwithapoptosis.Somecell ofactionoftheseosmiumarenecomplexes. sectionsappearnonstressedandjustdisplaymorecontrast Hydrophobicity, Cancer Cell Activity, and Cell Uptake. compared to the control cells (Figure 9A,B). Other cells ThelogPvalueisameasureforhydrophobicityandhasbeen appear distressed showing nuclear and cellular membranes investigated as a factor relevant to anticancer activity of withabnormalappearances(Figure9C,DandFigureS5). metal-based drugs for many years. For several classes Cells treated with 20 μM complex 3 appear much more of metalloanticancer complexes a correlation between in- distressedthanthecellstreatedwith5μM3,asdemonstrated creasedhydrophobicityandincreasedcytotoxicactivityhas byabnormal-lookingcellandnuclearmembranesforallcells beenreported.9,27-29SeveralstudiesreportonPtIIandPtIV (Figure 10A-D and Figure S6A-C). Many of the cells compounds,relatinghydrophobicitywithcelluptake.30-32 appeartobeundergoingapoptosis.Also,thesecellsdisplay Inthisstudy,thehydrophobicity,cancercellactivity,and evenmorecontrastcomparedtothecontrolcellsandthecells celluptakecorrelatedsignificantly,followingtheorder4∼3 treated with 5 μM 3. In particular, the osmium complex >2>1. Compound 1 was the least hydrophobic, the least Article JournalofMedicinalChemistry,2010,Vol.53,No.2 845 cytotoxic, and the least taken up by the cells, whereas while in the cytoplasm, the complexes exist as equilibrium compounds 4 and 3 displayed the highest hydrophobicity, mixturesoftheirchloridoandaquaforms.Theaquaspecies werethemostcytotoxic,andwerethemosttakenupbythe mayreactwithproteinsandorganellesinthecytoplasmand, cells.ThesedatasuggestthatintheovarianA2780cancercell being positively charged and lipophilic, may target mito- line,thelogPvaluecanbeusedtopredictthecytotoxicityfor chondria. On the other hand, the chlorido species, being thisclassofosmiumcompound.Thesedataalsoshowthat relativelyunreactive,areabletocrossthenuclearmembrane using more extended coordinated arenes such as biphenyl and then hydrolyze and react with the negatively charged andtetrahydroanthraceneresultsinincreasedhydrophobi- DNA. cityleadingtohighercellularuptakeandhighercytotoxicity. Compound 1 exhibits a different behavior: at chloride Inaddition,itislikelythatthecoordinatedarenesarealso concentrations of 4 and 22.7 mM, 1 is completely hydro- involvedininteractionswithpotentialbiologicaltargets.The lyzed, and at high NaCl concentrations of 100 mM about significantly lower cytotoxicity of the unsubstituted arene 50%ispresentinitschlorido“prodrug”form.Thepresence (benzene) compound 1 compared to the substituted arene of the reactive aqua species of 1 at relatively high NaCl complexes2-4indicatesthatthesubstituentsonthearenein concentrations may partly explain the reduced cell uptake these complexes do play a part in the mechanism of their andmoremoderatecytotoxicityofthiscomplexcomparedto cytotoxicity.Thiscouldberationalizedinpart,forexample, 2,3,and4.Thereactiveaquaadductof1maybedeactivated bytheintercalationoftheextendedarenesin3and4between by biomolecules on its way to its target either in blood DNAbaseswhenthecomplexesbindtothemajorgrooveof plasmaorinthecytoplasm. DNA, as is also observed for the analogous ruthenium(II) Distribution of the Osmium Complexes in Cell Fractions. complexes.33,34 In the case of complex 2, the arene (p- The uptake of the four complexes into the different cell cymene) may perturb DNA structure in the major groove fractionsfollowsthesametrendaswasobservedforwhole viastericinteractions(i.e.,uponbindingof2toanucleobase, cell uptake (i.e., the total amount of osmium in the frac- themethylandisopropylsubstituentsontheareneringmay tionsfollowstheorder3∼4>2>1).Importantly,thereisa causeadditionaldistortionstothestructureofDNA). correlationbetweenthenucleusuptakeandcytotoxicityof Time-dependent cell uptake experiments with complex 3 thecomplexes,bothofwhichfollowtheorder3>4>2>1, showedthatthemaximumcelluptakeisreachedafter12h, suggesting that penetrating the nucleus and binding to afterwhichtimeadecreaseinosmiumconcentrationandcell nuclear DNA might make a major contribution to the numberisobserved(Figure5).Thisshowsthattheoptimum mechanismofcytotoxicity.Thedistributionoftheosmium timeforinductionofthedeathofA2780ovariancancercells overtheseveralisolatedcellfractionsisverysimilarforall isbetween12and48h.Adecreaseinosmiumconcentration four osmium compounds; with 11-17% of the total ofabout55%aftermeasuringtheosmiumcontent24hafter of the osmium in the different fractions reaching the cell theendofa24hexposureto3indicatesthatosmiumefflux nucleus, 64-79% was found in the cytosolic fractions and fromthecellscanoccur. alsosignificantamountsofosmiumwerefoundintheouter AqueousReactivity.Improvedunderstandingoftheaqu- cell membrane and cytoskeleton fractions, in particular eouschemistryoforganometalliccomplexesunderbiologi- for compounds 2 and 3 (Table 5, Figure 6). This indicates cally relevant conditions should aid the rational design of that once in the cell, the osmium complexes distribute metalanticancercomplexes. over the cell independent of the nature of the arene or The hydrolysis rates for compounds 1, 2, and 4 were theirhydrophobicity.Significantamountsofosmiumreach measuredby1HNMRat288K.Allcomplexeshydrolyzed thecellnucleusforallfourcompounds.Inparticular,13% relativelyrapidlywithhalf-lives(t )between0.39and0.98 of the overall osmium uptake for compound 1 reaches 1/2 h.Thereappearstobenolinkbetweenthehydrolysisratesof the nucleus, displaying a similar distribution as com- thesearenecompoundsandtheirbiologicalactivity.Further- plexes 2-4. This is surprising because its aqueous che- more,increasingthetemperaturetothebiologicallyrelevant mistryissignificantlydifferentfromthatof2-4;thereactive temperature of 310K would increase the rate by about 8- aquaspeciesformsevenatrelativelyhighchlorideconcen- fold.18Thissuggeststhatat310K,allfourcomplexeswould trations. hydrolyzewithinafewminutestoformthereactivecationic DNAAdductFormation.Osmium(II)arenecomplexesare aqua species in low-chloride intracellular compartments. believedtointeractwithDNAinasimilarmannerastheir However, at the higher chloride concentration of 100 mM ruthenium analogues, i.e., binding to N7 of guanine in (typical of blood plasma), compounds 2-4 are 80-100% combinationwithH-bonding(ifthereisaneighboringNH present at equilibrium as the intact chloro (“prodrug”) on an Os ligand) and noncovalent arene intercalation into species. At a lower chloride concentration resembling that DNA.26 With the present data it is possible to examine inthecellcytoplasm(22.7mM),44-53%ofcomplexes2-4 whether thereisarelationshipbetweentheextentofDNA is present as the active hydrolyzed form, and at the lowest bindingandthecytotoxicityofthecomplexes. chloride concentration of 4 mM, close to that of the cell TheamountofosmiumfoundontheDNAofA2780cells nucleus,complexes2-4are56-81%presentasthereactive followstheorder;4.3>1>2anddoesnotcorrelatewith aqua species (Table 3). These data indicate that in the cell cytotoxic potency or hydrophobicity (logP) or cellular nucleus, complexes 2-4 might be selectively activated uptake (Table 5, Figure 3). The extent of osmium binding through hydrolysis as a mode of activation toward DNA of 1-4 to DNA, which amounts to 1.0-4.9% of the total binding, while outside the cell and in particular in blood osmiumtakenupbythecells,appearstobelow.However, plasma, complexes 2-4 are predominately present as their thisisasimilarleveltothatreportedforcellularplatination less reactive intact chlorido species. From this it can be ofDNAbycisplatin(∼1%).35-37Therefore,itisreasonable concludedthatthecomplexesaretakenupbycellspredomi- toassumethatDNAisapotentialtargetforthesecytotoxic nately in their neutral chlorido forms, their cell uptake osmiumcomplexes.Althoughcomplex4exhibitsthehighest dominatedbythehydrophobicityofthecoordinatedarenes, level of DNA binding, it is not the most potent complex 846 JournalofMedicinalChemistry,2010,Vol.53,No.2 vanRijtetal. (Table1).Thesedatasuggestnotonlythatitisimportantfor complexes that irreversibly inhibit mitochondrial human osmium to reach the DNA but also that the nature of the glutathione reductase (hGR) and thioredoxin reductase lesion on DNA (including structural distortions) is an im- (hTrxR).10,44 Another example is the structurally related portant factor in determining cytotoxic activity. However, RuII phosphine compound [Ru(η6-p-cymene)Cl (pta)]. Its 2 wecannotruleoutthepossibilitythatnuclearDNAmaynot cytotoxicityisthoughttobemediatedbymitochondrialand betheonlytargetforthesecomplexes. JNK-p53pathways.45 Distribution of Osmium Using TEM. TEM images of In previous work we have shown that osmium arene ovariancellsexposedtocomplex3for12hshowmorpho- complexesrelatedtothosestudiedherecanbindtoseveral logical changes to the cells that are commonly associated nucleobases,withahighaffinityforguanine.Wehaveshown with cell apoptosis.38 These include cell contraction that these complexes bind to calf thymus DNA and cause (Figure 7A, Figure S2A,B), dynamic membrane blebbing DNA unwinding (instead of DNA bending seen for ruthe- (Figure 7B, Figure S2C,D), and DNA fragmentation nium arene complexes).18,20,26,46,47 Our results support the (Figure7C,FigureS2E,F).Also,theformationofapoptotic possibilitythatpermeationoftheseosmiumcomplexesinto bodies was observed in several images, marking the final mitochondriaandthecellnucleusmaybefollowedbytheir stage of cell apoptosis (Figure 7D, Figure S3). We can binding to and distortion of the mitochondrial or nuclear therefore conclude that osmium complex 3 and probably DNA, leading tocell apoptosiswhich causesthe morphol- theotherstructurallysimilarosmiumcomplexesinducecell ogicalchangesobservedintheTEMimagesofcellsections. deathviaapoptoticpathwaysratherthannecrosis. Althoughitcannotbeconfirmedwhetherbindingofosmium The micrographs from cells exposed to 5 μM complex 3 tomitochondrialornuclearDNAisthemajorpathwayto illustrate that 3 accumulates mainly in the mitochondria, apoptosis, the images of cell sections provide the first evi- nucleolus,andnuclearmembrane(Figure9andFigureS5). dence that mitochondria may be involved for these and Several images show apoptotic cells and some cells appear relatedosmiumarenecomplexes. distressed, as evidenced by the abnormal looking nucleus and cellular membranes (Figure 9C,D and Figure S5). All Conclusions cellsexposedto20μMcomplex3(6.25timesitsIC value) 50 appeareddistressedand/orapoptotic.Also,inthesecellsthe This study shows that the hydrophobicity (logP), cancer accumulation of 3 in the mitochondria, nucleolus, and cellactivity,andcelluptakeforfourosmiumarenecomplexes nuclear membrane becomes even more clearly visible are significantly correlated. This suggests that logP values (Figure 10A-D and Figure S6A-D). In particular the maybeusefulindicatorsofA2780cancercellcytotoxicityfor mitochondriaareheavilystainedandtheinternalstructures this class of OsII arene complexes. Furthermore, this work ofthemitochondriabecomeclearlyvisible.Inaddition,the shows that the arene ligand plays an important part in the morphology of some of the mitochondria is very different cytotoxicity; the presence of substituted arenes not only from that observed for control cells and cells exposed to resulted in higher hydrophobicity and increased uptake by 5 μM 3; condensed cristae and swelling are observed the cell but also seemed to play an important part in the (Figure 10D-F and Figure S6D-F). The intensely dark- mechanismofcytotoxicityofthecomplexes. coloredmitochondriasuggestthatosmiumcomplex3accu- From the TEM images and from the cell fractionation mulates inside the cristae as well as in the inner and outer studies it appears that the osmium compounds accumulate membranes of the mitochondria. This suggests that mito- in cellular, mitochondrial, and nuclear membranes in signi- chondriamaybeatargetfortheseosmiumarenecomplexes ficant amounts. The TEM images of cell sections also re- or that the swelling and morphological changes of the vealed morphological changes associated with apoptosis mitochondriaariseasaconsequenceofapoptosis.Although in A2780 cells exposed to [(η6-biphenyl)Os(4-methyl- littleisknownabouttherelationshipbetweenalterationsin picolinate)Cl](3)excludingnecrosisaspossiblecauseforcell mitochondrialmorphologyandactivationofthemitochon- death for this complex and likely for structurally similar drial apoptotic pathway, a few studies have associated osmiumcomplexes. changes in mitochondrial morphology with apoptosis in- There is a correlation between nucleus uptake and cyto- ducedinculturedcellsbydifferentapoptoticstimuli.These toxicityofthecompounds,suggestingthatnucleusuptakeand changes have mainly been described as disorganization, binding to nuclear DNA may be the major pathway for condensation,orswelling.39-43 cytotoxicity. However, the levels of osmium binding to Fromthecellfractionationstudiesitwasshownthatabout DNAwererelativelylowanddonotcorrelatewithcytotoxic three-quartertoone-thirdoftheosmiumthatistakenupby potency.Thisincombinationwiththemitochondrialswelling cellsispresentinthecytosolicfractionafter24hofexposure andmorphologychangesobservedbyTEMimagingsuggests to compounds 1-4, indicating that the complexes may be thatthenatureoftheDNAlesionmaybeofimportanceand interacting with organelles and biomolecules present in alsoraisesthepossibilitythatmitochondrialapoptoticpath- the cytosol. In addition, the osmium complexes appear to waysareinvolved. haveaffinityforthecellularmembranes.Thisisdemonstrated by the dark-colored nuclear membrane in the TEM images ExperimentalSection andbythecell fractionationstudies,inwhichlargeamounts Materials.1,4-Dihydrobiphenylandthedimers[(η6-p-cym)- ofosmiumweredetectedinthemembrane/cytoskeletonfrac- OsCl ] , [(η6-bip)OsCl ] , [(η6-bz)OsCl ] , and [(η6-THA)Os- tions, especially for complexes 2 and 3. Mitochondria may 22 22 22 Cl ] werepreparedusingpreviouslyreportedprocedures.23,24 represent potential targets, and the abundance of osmium in 22 OsCl 3nH O was purchased from Alfa Aesar. All deuterated mitochondrialmembranesmayexplainwhysomuchosmium 3 2 solvents and cisplatin were obtained from Sigma Aldrich. remainsinthecytosol. Methanol was distilled over magnesium/iodine prior to use. Other metal-based anticancer complexes known to Complexes 1-4 were synthesized from the dimeric precur- target mitochondria include several gold(I) and gold(III) sors [(η6-p-cym)OsCl ] , [(η6-bip)OsCl ] , [(η6-bz)OsCl ] , and 22 22 22 Article JournalofMedicinalChemistry,2010,Vol.53,No.2 847 [(η6-THA)OsCl ] using procedures similar to those reported NMR spectra were taken after various intervals using the 22 previously for other half-sandwich OsII arene complexes.48,49 presaturationmethodforwatersurpression.Therateofhydro- The4-methylpicolinicacidligandandcomplex3,[(η6-bip)Os- lysis was determined by fitting plots of concentrations (4-methyl-picolinate)Cl],weresynthesizedandcharacterizedas (determined from 1H NMR peak integrals) versus time to a previouslydescribed.20Thepuritiesofcompounds1,2,and4 pseudo-first-order equation using ORIGIN, version 7.5 werealldeterminedtobeg95%byelementalanalysisandare (MicrocalSoftwareLtd.). reportedintheSupportingInformation.Inaddition,1HNMR Theeffectsofvaryingconcentrationsofchlorideontheextent andESI-MSdataforcompounds1,2,and4canalsobefoundin of hydrolysis of the complexes at equilibrium were investi- theSupportingInformation. gated by preparing aqueous solutions of 1-4 (1 mM) in 100, Preparation of the Complexes. [(η6-Benzene)Os(4-methyl- 22.7,and4mMNaClinD O,recordingspectraafterenough 2 picolinate)Cl] (1). A solution of [(η6-bz)OsCl ] (75 mg, 0.11 timewasleftforequilibration(atleast4timestheirrespective 22 mmol)indryanddegassedMeOH(10mL)wasrefluxedunder half-lives). argonfor1hbeforeaddingasolutionofsodiummethoxide(2.2 CellCulture.TheA2780ovariancelllinewasobtainedfrom molequiv,13mg)and4-methyl-picolinicacid(2.2molequiv, the ECACC (European Collection of Animal Cell Culture, 39 mg) in 5 mL of dry and degassed MeOH. The resulting Salisbury, U.K.). The cells were maintained in RPMI 1640 mixture was left to reflux mildly for 16 h, at which time a mediawhichwassupplementedwith10%fetalcalfserum,1% crystalline brown precipitate had formed. The brown powder L-glutamine, and 1% penicillin/streptomycin. All cells were wasrecoveredbyfiltrationandwasair-driedtogiveafinalyield grownat310Kinahumidifiedatmospherecontaining5%CO 2 . of30.7mg(32%). In Vitro Growth Inhibition Assay. After plating, human [(η6-p-Cymene)Os(4-methyl-picolinate)Cl] (2). A solution of ovarian A2780 cancer cells were treated with OsII complexes [(η6-p-cym)OsCl ] (70 mg, 0.103 mmol) in dry and degassed on day 3 at concentrations ranging from 0.1 to 100 μM. 22 MeOH(10mL)wasrefluxedunderargonfor1hbeforeaddinga SolutionsoftheOsIIcomplexesweremadeupin0.125%(v/v) solution of sodium methoxide (2.6 mol equiv, 14.5 mg) and DMSO to assist dissolution (0.03% final concentration of 4-methyl-picolinicacid(2.6molequiv,36.7mg)in5mLofdry DMSOperwellinthe96-wellplate).Cellswereexposedtothe and degassed MeOH. The resulting mixture was leftto reflux complexesfor24h,washed,suppliedwithfreshmedium,and mildlyfor3handwasfiltered,andthesolventwasreducedona allowedtogrowfor3doublingtimes(72h),andthentheprotein rotaryevaporatoruntilaprecipitatebegantoformandwasleft contentwasmeasured(proportionaltocellsurvival)usingthe standingat278K.Theyellowpowderwasrecoveredbyfiltra- sulforhodamineB(SRB)assay.50 tionandwasair-driedtogiveafinalyieldof46.8mg(46%). Drug Uptake, DNA Adduct Formation, and Cell Fractiona- [(η6-Tetrahydroanthracene)Os(4-methyl-picolinate)Cl] (4). tion.A2780cellswereplatedatadensityof5(cid:2)106cells/100mm Synthesis was as for 2 using [(η6-THA)OsCl ] (67 mg, 0.076 Petri dishin 12mL of culture medium onday 1 (three dishes 22 mmol), sodium methoxide (2.6 mol equiv, 29 mg), and werepreparedpercompoundtestedandthreeuntreatedcontrol 4-methyl-picolinicacid(2.6molequiv,10mg).Yield:25.6mg dishesintwoindependentseparateexperiments).Onday2cells (31%). wereexposedtotheOsIIarenecomplexes.Stocksolutions(0.5mM) Instrumentation and Methods. Nuclear Magnetic Resonance ofosmiumcompoundsweremadeupfreshin5%DMSOand (NMR)Spectroscopy.1HNMRspectrawereacquiredin5mm salinebeforebeingdilutedinmediatogiveafinalconcentration NMR tubes at 298 K (unless stated otherwise) on a Bruker of 5 μM. After 24 h of drug exposure, the drug-containing DMX500(1H=500.13MHz)spectrometer.1HNMRchemical medium was removed. The cells were washed with PBS and shiftswereinternallyreferencedto(CHD )(CD )SO(2.50ppm) trypsinized,andthecellsuspensionwascounted.One-thirdof 2 3 forDMSO-d andtoCHD OD(3.34ppm)formethanol-d thecellswerecentrifuged,washedwithPBS,andstoredat253K. 6 2 4. ElementalAnalysis.Carbon,hydrogen,andnitrogen(CHN) Another third of the samples was used for DNA extraction elemental analyses were carried out on a CE-440 elemental usingtheNucleongenomicDNAextractionkit(GEhealthcare, analyzerbyExeterAnalytical(UK)Ltd. Amersham, U.K.) (BACC-1 protocol) and the last third for ICP-MSInstrumentationandCalibration.AllICP-MSana- cytosol, nucleus, membrane/cytoskeleton fractionation using lyses were carried out on an Agilent Technologies 7500 series a cell fractionation kit (Biovision, Mountain View, CA) ICP-MSinstrument.ThesettingsforICP-MSarelistedinTable (fractionPREP). All samples were stored at 253 K until ICP- S1.ThewaterusedforICP-MSanalysiswasdoublydeionized MSanalysisforosmiumcontent. (DDW)usingaUSFElgaUHQwaterdeionizer.Theosmium Forthetime-dependentcelluptakeexperiments,A2780cells Specpureplasmastandard(AlfaAesar,1000ppmin5%HCl) wereplatedatadensityof5(cid:2)106cells/100mmPetridishin wasdilutedwithDDWto20ppm.Thestandardsforcalibration 12mLofculturemediumonday1(threedisheswereprepared were freshly prepared bydiluting this stocksolution with 3% per time point and three untreated control dishes). On day 2, HNO inDDW.Theconcentrationsusedwere100,60,20,10,5, cells were exposed to complex 3 at 5 μM (prepared as stated 3 4,2,1,0.4,and0.1ppb.Withtheinstrumentationsettingslisted previously)at1,2,4,8,12,and24htimeintervals.Afterthe in Table S1, the detection limit was typically 9 ppt (for 10 indicated exposure times, cells were trypsinized and the cell standards),andsensitivitywas270900189Osioncountsfor100 suspension was counted. The cells were centrifuged, washed ppbofOsstandardinnogasmode. with PBS, and stored at 253 K until ICP-MS analysis for ESI-MS.Electrosprayionizationmassspectrawereobtained osmiumcontent. either on a Bruker Esquire 2000 spectrometer or a Bruker Toestablishtheosmiumuptakeaftera24hexposureand24h MicroTOFspectrometer.Sampleswerepreparedineitherwater after the end of the exposure to complex 3, A2780 cells were oramethanol/watermixture,andtheconevoltageandsource platedatadensityof5(cid:2)106cells/100mmPetridishin12mLof temperature varied depending on the sample. Data were pro- culturemediumonday1(threedisheswerepreparedforthe24h cessedusingDataAnalysis3.3(BrukerDaltonics). exposure,threedishesfor24hofexposureand24hofrecovery AqueousReactivity.Thekineticsofhydrolysisforcomplexes indrug-freemedia,andthreeuntreatedcontroldishes).Onday2, 1-4werefollowedby1HNMRatdifferenttemperatures.For cells were exposed to complex 3 at 5 μM, prepared as stated this,solutionsofthecomplexeswithafinalconcentrationof0.8 previously.After24hofexposureto3,cellsweretrypsinisedand mMin5%MeOD-d /95%D O(v/v)werepreparedbydissolu- the cell suspension was counted. The cells were centrifuged, 4 2 tionofthecomplexesinMeOD-d followedbyrapiddilution washedwithPBS,andstoredat253KuntilICP-MSanalysisfor 4 using D O with a pH* (pH meter reading without correction osmiumcontent.Forthreedishes,thedrugcontainingmedium 2 for effects of D on glass electrode) of ∼2 (acidified with was removed, cells were washed with PBS, and fresh medium HNO ) so that the aqua ligand was not deprotonated. 1H was supplied. After a 24 h recovery period, the medium was 3 848 JournalofMedicinalChemistry,2010,Vol.53,No.2 vanRijtetal. removed,cellswerewashedwithPBSandtrypsinized,andthe themtoexchangeregularlythemostrecentideasinthefieldof cellsuspensionwascounted.Thecellswerecentrifuged,washed anticancermetallodrugswithseveralEuropeancolleagues. with PBS, and stored at 253 K until ICP-MS analysis for osmiumcontent. Supporting Information Available: Spectroscopic data and ICP-MSAnalysis.Thewholecellpellets,DNAsamples,and ESI-MS and CHN analysis data for complexes 1, 2, and 4; cytosol/nucleus/membranefractionationsamplesweredigested instrumentalsettingsforICP-MS(TableS1);timedependence asdescribedbelow.Tothecellpellets0.8mLandtotheextracts forformationoftheaquacomplexesof1,2,and4(FigureS1); 0.4 mL of freshly distilled 72% HNO were added, and the 3 TEMimagesofA2780cellsexposedtocomplex3,showingthe samples were transferred into Wheaton V-Vials (Sigma different stages of cell apoptosis (Figures S2 and S3); TEM Aldrich).Thevialswereheatedinanovenat373Kfor16hto images of A2780 cells only stained with 2% uranyl acetate fully digest the samples and allowed to cool, and then each (controls)(FigureS4);TEMimagesofA2780cellsexposedto samplewastransferredtoaFalcontube.Thevialswerewashed 5 μM complex 3 (Figure S5); TEM images of A2780 cells withdoublydeionizedwater(DDW)andthesamplesdiluted10 exposed to 20 μM complex 3 (Figure S6). This material is timeswithDDWtoobtain7.2%HNO samplesolutions. 3 availablefreeofchargeviatheInternetathttp://pubs.acs.org. Determination of Partition Coefficient, logP. Octanol-satu- rated water (OSW) and water-saturated octanol (WSO) were prepared using analytical grade octanol (Sigma) and 0.3 M References aqueousNaClsolution.Aliquotsofstocksolutionsofosmium (1) Tredan, O.; Galmarini, C. M.; Patel, K.; Tannock, I. F. Drug complexesinOSWwereaddedtoequalvolumesofWSOand resistanceandthesolidtumormicroenvironment.J.Natl.Cancer shakeninanIKAVibraxVXCbasicshakerfor4hat500g/min Inst.2007,99,1441–1454. afterpartition.Theaqueousandoctanollayerswerecarefully (2) Minchinton, A. I.; Tannock, I. F. Drug penetration in solid tumours.Nat.Rev.Cancer2006,6,583–592. separatedintotesttubesforosmiumanalysis.Aqueoussamples (3) Ishida, S.; Lee, J.; Thiele, D. J.; Herskowitz, I. Uptake of the beforeandafterpartitioningweredilutedwith3%HNO tothe 3 anticancerdrugcisplatinmediatedbythecoppertransporterCtr1 appropriate range for analysis by ICP-MS calibrated with in yeast and mammals. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, aqueous standards (Os, 1-100 ppb). These procedures were 14298–14302. carriedoutatambienttemperature(∼298K).Partitioncoeffi- (4) Wang,D.;Lippard,S.J.Cellularprocessingofplatinumanticancer cientsofOsIIarenecomplexeswerecalculatedusingtheequa- drugs.Nat.Rev.DrugDiscovery2005,4,307–320. (5) Borst, P.; Evers, R.; Kool, M.; Wijnholds, J. A family of drug tionlogP =log([Os] /[Os] ). oct oct aq transporters:themultidrugresistance-associatedproteins.J.Natl. TEM Sample Preparations. A2780 cells were plated at a CancerInst.2000,92,1295–1302. densityof5(cid:2)106cells/100mmPetridishin12mLofculture (6) Safaei, R.; Howell, S. B. Copper transporters regulate the medium(twodisheswerepreparedforexposuretocompound3 cellularpharmacologyandsensitivitytoPtdrugs.Crit.Rev.Oncol. and two dishes for untreated control plates). After 24 h, Hematol.2005,53,13–23. (7) Klein,A.V.;Hambley,T.W.Platinumdrugdistributionincancer untreated cells received fresh media while treated cells were cellsandtumors.Chem.Rev.2009,109,4911–4920. exposedtocomplex3at20and5μM.Stocksolutions(0.5mM) (8) Rabik,C.A.;Dolan,M.E.Molecularmechanismsofresistance ofosmiumcompoundsweremadeupfreshin5%DMSOand andtoxicityassociatedwithplatinatingagents.CancerTreat.Rev. salinebeforebeingdilutedinmediatogivefinalconcentrations 2007,33,9–23. of20and5μM.After12hofdrugexposure,themediumwas (9) Loh,S.Y.;Mistry,P.;Kelland,L.R.;Abel,G.;Harrap,K.R. Reduced drug accumulation as a major mechanism of acquired removed and the cells were washed with PBS. The cells were resistance to cisplatin in a human ovarian carcinoma cell line: fixedwith2%gluteraldehyde(AgarScientific,Cambridge,U.K.) circumventionstudiesusingnovelplatinum(II)and(IV)ammine/ incacodylatebufferatpH7.6(AgarScientific,Cambridge,U.K.) aminecomplexes.Br.J.Cancer1992,66,1109–1115. for 1 h at ambient temperature (agitating on a shaker). The (10) Berners-Price, S. J.; Filipovska, A. The design of gold-based, plates wererinsedfor10min withcacodylatebuffer,andthis mitochondria-targeted chemotherapeutics. Aust. J. Chem. 2008, 61,661–668. was repeated five times. The cells were scraped with a cell (11) Dyson,P.J.;Sava,G.Metal-basedantitumordrugsinthepost scraper and transferred in 0.5 mL of buffer into a 0.5 mL genomicera.DaltonTrans.2006,1929–1933. Eppendorf tube. The samples were stained with 2% uranyl (12) Jakupec, M. A.; Galanski, M.; Arion, V. B.; Hartinger, C. G.; acetate in maleate buffer (Agar Scientific, Cambridge, U.K.) Keppler,B.K.Antitumormetalcompounds:morethanthemeand variations.DaltonTrans.2008,183–194. for1hatambienttemperature.Cellswerethendehydratedin (13) Melchart, M.; Sadler, P. J. Ruthenium Arene Anticancer Com- graded series of ethanol with a final change of 100% ethanol plexes. In Bioorganometallics; Jaouen, G., Ed.; Wiley-VCH: overnight. Prior to embedding, the cells were infiltrated with Weinheim,Germany,2006;pp39-64. 100% propylene oxide for 1 h followed by a 1:1 mixture of (14) Reedijk,J.Metal-ligandexchangekineticsinplatinumandruthe- propylene oxide and Embed 812 resin (Agar Scientific, Cam- niumcomplexes.Significanceforeffectivenessasanticancerdrugs. PlatinumMet.Rev.2008,52,2–11. bridge, U.K.) for 6 h. This was then replaced with several (15) Bruijnincx,P.C.A.;Sadler,P.J.Newtrendsformetalcomplexes changes of 100% resin over an 18 h period before curing for withanticanceractivity.Curr.Opin.Chem.Biol.2008,12,197–206. 24hat333K.BlocksweretrimmedandsectionedonaLeica (16) van Rijt, S. H.; Sadler, P. J. Current applications and future Ultracut E ultramicrotome (Leica Microsystems, Milton potentialforbioinorganicchemistryinthedevelopmentofanti- Keynes,U.K.),and70-80nmsectionswerecollectedon400- cancerdrugs.DrugDiscoveryToday2009,14,1089–1097. (17) Dorcier, A.; Ang, W. H.; Bolano, S.; Gonsalvi, L.; Juillerat- mesh copper grids and stained with uranyl acetate. The Jeannerat, L.; Laurenczy, G.; Peruzzini, M.; Phillips, A. D.; sections were imaged on a JEOL 1200EXII TEM with a 1K Zanobini, F.; Dyson, P. J. In vitro evaluation of rhodium and Gatancamera. osmiumRAPTAanalogues:thecasefororganometallicanticancer drugsnotbasedonruthenium. Organometallics2006,25,4090– 4096. Acknowledgment. WethankDr.CarolEvered(University (18) Peacock,A.F.A.;Parsons,S.;Sadler,P.J.Tuningthehydrolytic ofWarwick,U.K.)forhelpwiththepreparationandsection- aqueouschemistryofosmiumarenecomplexeswithN,O-chelating ligandstoachievecancercellcytotoxicity.J.Am.Chem.Soc.2007, ingoftheTEMsamplesandDr.LijiangSong(Universityof 129,3348–3357. Warwick, U.K.) for help with ICP-MS. This research was (19) Schmid,W.F.;John,R.O.;Arion,V.B.;Jakupec,M.A.;Keppler, supported by the EC (Marie Curie Fellowship for A.M.), B.K.Highlyantiproliferativeruthenium(II)andosmium(II)arene MRC(GrantG0701062),UniversityofWarwickandScience complexes with paullone-derived ligands. Organometallics 2007, 26,6643–6652. City (ERDF/AWM). The authors also acknowledge their (20) vanRijt,S.H.;Peacock,A.F.A.;Johnstone,R.D.L.;Parsons, participation in the EU COST Action D39 which enabled S.; Sadler, P. J. Organometallic osmium(II) arene anticancer Article JournalofMedicinalChemistry,2010,Vol.53,No.2 849 complexescontainingpicolinatederivatives.Inorg.Chem.2009,48, (36) Jung, Y.; Lippard, S. J. Direct cellular responses to platinum- 1753–1762. inducedDNAdamage.Chem.Rev.2007,107,1387–1407. (21) Peacock,A.F.A.;Sadler,P.J.Medicinalorganometallicchemi- (37) Yu,F.;Megyesi,J.;Price,P.M.Cytoplasmicinitiationofcisplatin stry:designingmetalarenecomplexesasanticanceragents.Chem. cytotoxicity.Am.J.Physiol.2008,295,F44–F52. AsianJ.2008,3,1890–1899. (38) Kerr, J. F.; Wyllie, A. H.; Currie, A. R. Apoptosis: a basic (22) van Rijt, S. H.; Hebden, A. J.; Amaresekera, T.; Deeth, R. J.; biological phenomenon with wide-ranging implications in tissue Clarkson,G.J.;Parsons,S.;McGowan,P.C.;Sadler,P.J.Amide kinetics.Br.J.Cancer1972,26,239–257. linkageisomerismasanactivityswitchfororganometallicosmium (39) Gottlieb, E.; Armour, S. M.; Harris, M. H.; Thompson, C. B. and ruthenium anticancer complexes. J. Med. Chem. 2009, 52, Mitochondrialmembranepotentialregulatesmatrixconfiguration 7753–7764. andcytochromecreleaseduringapoptosis.CellDeathDiffer.2003, (23) Stahl, S.; Werner, H. A new family of (arene)osmium(0) and 10,709–717. (arene)osmium(II)complexes.Organometallics1990,9,1876–1881. (40) Kohler,H.R.;Dhein,J.;Alberti,G.;Krammer,P.H.Ultrastruc- (24) Peacock,A.F.A.;Habtemariam,A.;Fernandez,R.;Walland,V.; turalanalysisofapoptosisinducedbythemonoclonalantibody Fabbiani,F.P.A.;Parsons,S.;Aird,R.E.;Jodrell,D.I.;Sadler, anti-APO-1onalymphoblastoidBcellline.Ultrastruct.Pathol. P.J.Tuningthereactivityofosmium(II)andruthenium(II)arene 1990,14,513–518. complexesunderphysiologicalconditions.J.Am.Chem.Soc.2006, (41) Papadimitriou, J. C.; Drachenberg, C. B.; Shin, M. L.; Trump, 128,1739–1748. B.F.Ultrastructuralstudiesofcomplementmediatedcelldeath:a (25) Wang,F.;Chen,H.M.;Parsons,S.;Oswald,L.D.H.;Davidson, biologicalreactionmodeltoplasmamembraneinjury.Virchows J.E.;Sadler,P.J.Kineticsofaquationandanationofruthenium- Arch.1994,424,677–685. (II) arene anticancercomplexes,acidity andX-ray structuresof (42) Scorrano, L.; Ashiya, M.; Buttle, K.; Weiler, S.; Oakes, S. A.; aquaadducts.Chem.;Eur.J.2003,9,5810–5820. Mannella,C.A.;Korsmeyer,S.J.Adistinctpathwayremodels (26) Kostrhunova,H.;Florian,J.;Novakova,O.;Peacock,A.F.A.; mitochondrial cristae and mobilizes cytochrome c during apo- Sadler, P. J.; Brabec, V. DNA interactions of monofunctional ptosis.Dev.Cell2002,2,55–67. organometallicosmium(II)antitumorcomplexesincell-freemedia. (43) Yasugi,E.;Uemura,I.;Kumagai,T.;Nishikawa,Y.;Yasugi,S.; J.Med.Chem.2008,51,3635–3643. Yuo, A. Disruption of mitochondria is an early event during (27) Barbieri,R.QSARapproachtounderstandtheantitumoractivity dolichyl monophosphate-induced apoptosis in U937 cells. Zool. oforganotins.Inorg.Chim.Acta1992,191,253–259. Sci.2002,19,7–13. (28) Mendoza-Ferri, M.-G.; Hartinger, C. G.; Eichinger, R. E.; (44) Bindoli,A.;Rigobello,M.P.;Scutari,G.;Gabbiani,C.;Casini,A.; Stolyarova, N.; Severin, K.; Jakupec, M. A.; Nazarov, A. A.; Messori,L.Thioredoxinreductase:atargetforgoldcompounds Keppler, B. K. Influence of the spacer length on the in vitro actingaspotentialanticancerdrugs.Coord.Chem.Rev.2009,253, anticancer activity of dinuclear ruthenium-arene compounds. 1692–1707. Organometallics2008,27,2405–2407. (45) Chatterjee, S.; Kundu, S.; Bhattacharyya, A.; Hartinger, C. G.; (29) Song,R.;Park,S.Y.;Kim,Y.-S.;Kim,Y.;Kim,S.-J.;Ahn,B.T.; Dyson, P. J. The ruthenium(II)-arene compound RAPTA-C Sohn,Y.S.Synthesisandcytotoxicityofnewplatinum(IV)com- induces apoptosis in EAC cells through mitochondrial and plexesofmixedcarboxylates.J.Inorg.Biochem.2003,96,339–345. p53-JNKpathways.J.Biol.Inorg.Chem.2008,13,1149–1155. (30) Hall,M.D.;Amjadi,S.;Zhang,M.;Beale,P.J.;Hambley,T.W. (46) Peacock,A.F.A.;Habtemariam,A.;Moggach,S.A.;Prescimone, Themechanismofactionofplatinum(IV)complexesinovarian A.; Parsons, S.; Sadler, P. J. Chloro half-sandwich osmium(II) cancercelllines.J.Inorg.Biochem.2004,98,1614–1624. complexes: influence of chelated N,N-ligands on hydrolysis, (31) Oldfield,S.P.;Hall,M.D.;Platts,J.Acalculationoflipophilicity guaninebinding,andcytotoxicity.Inorg.Chem.2007,46,4049– ofalarge,diversedatasetofanticancerplatinumcomplexesandthe 4059. relationtocellularuptake.J.Med.Chem.2007,50,5227–5237. (47) Peacock,A.F.A.;Melchart,M.;Deeth,R.J.;Habtemariam,A.; (32) Platts,J.A.;Hibbs,D.E.;Hambley,T.W.;Hall,M.D.Calculation Parsons, S.; Sadler, P. J. Osmium(II) and ruthenium(II) arene ofthehydrophobicityofplatinumdrugs.J.Med.Chem.2001,44, maltolato complexes: rapid hydrolysis and nucleobase binding. 472–474. Chem.;Eur.J.2007,13,2601–2613. (33) Chen, H.; Parkinson, J. A.; Parsons, S.; Coxall, R. A.; Gould, (48) Morris, R. E.; Aird, R. E.; Murdoch, P. D.; Chen, H. M.; R. O.; Sadler, P. J. Organometallic ruthenium(II) diamine anti- Cummings,J.;Hughes,N.D.;Parsons,S.;Parkin,A.;Boyd,G.; cancer complexes: arene-nucleobase stacking and stereospecific Jodrell, D. I.; Sadler, P. J. Inhibition of cancer cell growth by hydrogen-bondinginguanineadducts.J.Am.Chem.Soc.2002, ruthenium(II) arene complexes. J. Med. Chem. 2001, 44, 3616– 124,3064–3082. 3621. (34) Aird,R.E.;Cummings,J.;Ritchie,A.A.;Muir,M.;Morris,R.E.; (49) Fernandez, R.; Melchart, M.; Habtemariam, A.; Parsons, S.; Chen,H.;Sadler,P.J.;Jodrell,D.I.Invitroandinvivoactivityand Sadler,P.L.Useofchelatingligandstotunethereactivesiteof cross resistance profiles of novel ruthenium (II) organometallic half-sandwichruthenium(II)-areneanticancercomplexes.Chem.; arenecomplexesinhumanovariancancer.Br.J.Cancer2002,86, Eur.J.2004,10,5173–5179. 1652–1657. (50) Skehan,P.;Storeng,R.;Scudiero,D.;Monks,A.;McMahon,J.; (35) Eastman,A.Activationofprogrammedcelldeathbyanticancer Vistica,D.;Warren,J.T.;Bokesch,H.;Kenney,S.;Boyd,M.R. agents:cisplatinasamodelsystem.CancerCells(1989)1990,2, Newcolorimetriccytotoxicityassayforanticancer-drugscreening. 275–280. J.Natl.CancerInst.1990,82,1107–1112.