Highly cytotoxic substitutionally inert rhodium(III) tris(chelate) complexes: DNA binding modes and biological impact on human cancer cells.

JournalofInorganicBiochemistry105(2011)991–999 ContentslistsavailableatScienceDirect Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio Highly cytotoxic substitutionally inert rhodium(III) tris(chelate) complexes: DNA binding modes and biological impact on human cancer cells Frauke Hackenberga, Luciano Oehningerb, Hamed Alborziniac, Suzan Canc, Igor Kitanovicc, Yvonne Geldmachera, Malte Kokoschkaa, Stefan Wölflc, Ingo Ottb, William S. Sheldricka,⁎ aFakultätfürChemieundBiochemie,Ruhr-UniversitätBochum,44780Bochum(Germany) bInstituteofMedicinalandPharmaceuticalChemistry,TechnischeUniversitätBraunschweig,Beethovenstrasse55,38106Braunschweig(Germany) cInstitutfürPharmazieundMolekulareBiotechnologie,Ruprecht-Karls-Universität,Heidelberg,ImNeuenheimerfeld364,69120Heidelberg(Germany) a r t i c l e i n f o a b s t r a c t Articlehistory: Theantiproliferativepropertiesandcellularimpactofnovelsubstitutionallyinertrhodium(III)complexesof Received8February2011 thetypes[Rh{(CH ) NCS } (pp)]Cl3–5(pp=5,6-Me phen,dpq,dppz)andOC-6-23-[Rh(2-S-py) (pp)]Cl6 3 2 22 2 2 Receivedinrevisedform14March2011 and7(2-S-py=pyridine-2-thiolate;pp=dpq,dppz)havebeeninvestigatedfortheadherenthumancancer Accepted13April2011 cell lines MCF-7 and HT-29 and for non-adherent Jurkat cells. Whereas CD and viscosity measurements Availableonline21April2011 indicatethatthepolypyridylligandsof4and5intercalateintoCTDNA,thisisnotthecasefortheanalogous pyridine-2-thiolatecomplexes6and7.Complexes3–7allexhibitahighantiproliferativeactivitytowards Keywords: MCF-7andHT-29cells,withIC valuesintherange0.055–0.285μM.Asestablishedbyonlinemonitoring Rhodium 50 Polypyridylligands with a cell-based sensor chip, the highly cytostatic complex 6 (IC 50 =0.059 and 0.078μM) invokes an DNAbinding immediate concentration-dependent reduction of MCF-7 cell respiration and a time-delayed decrease in Anticanceragents cellularimpedance,whichcanbeascribedtotheinductionofcelldeath.AnnexinV/PIassaysdemonstrated Cellmetabolism that 6 also has a pronounced antiproliferative activity towards Jurkat cells and that it invokes extensive Apoptosis apoptosisandhighconcentrationsofreactiveoxygenspeciesintheseleukemiacells.Theobservationofa dose-dependent inhibition of the oxygen consumption of isolated mice mitochondria indicates the involvementofanintrinsicmitochondrialpathwayinthisprocess. ©2011ElsevierInc.Allrightsreserved. 1.Introduction Althoughmanystudieshavebeendevotedtothemodeofinteraction oftris(chelate)complexeswithDNA,surprisinglyfewreportsoftheir Substitutionallyinertoctahedraltransition-metalcomplexescontain- possibleanticanceractivityhaveappearedandthesehaveconcentrated ing polypyridyl (pp) ligands have the ability to target specific base ondicationicruthenium(II)complexesofthetype[Ru(L1) (L2)]2+[5–8]. 2 sequencesinDNAandhave,asaresult,attractedconsiderableattentionas With the apparent exception of [Ru(bpy) (dppn)]2+containing the 2 potentialdiagnosticagents[1–3].Whilesimpletris(chelate)complexes very large dppn (benzo[i]dipyrido[3,2-a :2′,3′-c]phenazine) ligand, [M(pp) ]n+ (M=Ru, n=2; M=Rh, n=3) with three 2,2′-bipyridyl which exhibits in vitro IC values of 6.4±1.9 and 3.3±1.2μM [5] 3 50 (bpy)or1,10-phenanthroline(phen)ligandshavebeenshowntobe comparabletothoseofcisplatin(7.0±2.0and2.0±0.3μM)[9]towards groovebindersorpossiblepartialintercalators[1,2],strongintercalative thehumancancercelllinesHT-29andMCF-7,onlymoderatetolow bindinghasoftenbeenestablishedwhenalargerpolypyridylligandsuch activitywasestablished.Interestingly,atenfoldhigherantiproliferative aschrysi(5,6-chrysenquinonediimine)ordppz(dipyrido[2,3-a:2′,3′-c] activity(IC =0.64±0.08μM)towardsHT-29cellshasrecentlybeen 50 phenazine)ispresent.AstrikingexampleforspecificsiteDNAbindingis reportedforthekineticallyinertmonocationic rhodium(III)complex providedbythebulkycomplexΔ-[Rh(bpy) (chrysi)]3+whichinterca- [(η5-C Me )Rh(C H S)(dppz)]+[9]. This suggests that reducing the 2 5 5 6 5 lates into destabilized regions close to base pair mismatches and overall charge of substitutionally inert polypyridyl complexes may recognizesasinglemismatchina2725basepairplasmid[4]. possiblyhaveabeneficialeffectontheirantitumouractivity.Toestablish whetherthisisindeedthecase,wehavenowsynthesizedaseriesof monocationic rhodium(III) complexes of the type [Rh(L) (pp)]+ 2 containingtheanionicbidentateligands(L)dimethyldithiocarbamate ⁎ Correspondingauthor.Tel.:+492343228019;fax:+492343214420. (Scheme1)andpyridine-2-thiolate(Scheme2).Asthemechanismof E-mailaddress:William.sheldrick@rub.de(W.S.Sheldrick). actionofsuchmetallodrugsmustbebasedonnoncovalentinteractions, 0162-0134/$–seefrontmatter©2011ElsevierInc.Allrightsreserved. doi:10.1016/j.jinorgbio.2011.04.006 992 F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 Fig.1.DFTstructureofcomplex5. Scheme1.Structuresofcomplexes1–5. wehavealsostudiedtheirpossibleintercalativeDNAbindingmodes protons of dpq complex 6 (Fig. 3). An analogous pattern was andthebiologicalimpactofselectedcompoundsonhumancancercells. observedforthedppzprotonsofcomplex7.InadditiontotheOC-6- 23isomer,twosymmetricalligandarrangements,OC-6-13(StranstoS′) 2.Resultsanddiscussion andOC-6-33(SandS′transtoN(pp))arealsopossibleforcomplexesof thetype[Rh(2-S-py) (pp)]Cl.Forcefieldcalculationsindicatedthatthe 2 2.1.Synthesisandstructuresof1–7 OC-6-13isomershould,infact,beenergeticallymorefavorablethanthe isolatedOC-6-23isomerfor7.ThethirdOC-6-33isomerhasbothitsH6 Thedimethyldithiocarbamatocomplexes[Rh(Me NCS ) (pp)]Cl protons pointed above the dppz ligand and is energetically less 2 2 2 1–5 (pp=bpy, phen, 5,6-Me phen, dpq=dipyrido[3,2-d:2′,3′:f] favourable than the OC-6-23 arrangement. In contrast to 6 and 7, 2 quinoxaline,dppz)werepreparedinastepwisemannerfrommer, attempts to prepare the analogous OC-6-23 isomers for rhodium(III) cis-[RhCl (DMSO-κS) (DMSO-κO)] [10,11]. The precursor was first complexescontainingthesmallerpolypyridylligandsbpy,phenand5,6- 3 2 treated with the appropriate polypyridyl ligand to generate the Me 2 phenallledtomixturesofasymmetricalisomer(eitherOC-6-13or intermediate complexes [RhCl (DMSO)(pp)] [12], which were OC-6-33) and the asymmetrical OC-6-23 isomer. The ratio of the 3 subsequently refluxed with two equivalents of sodium dimethyl- symmetrical to the asymmetrical isomer decreased from 56:44 over dithiocarbamateinmethanoltoafford1–5,oncoolingthereaction 40:60 to 33:67 with increasing polypyridyl ligand size in the order solution to room temperature. A similar two-step procedure was bpybphenb5,6-Me 2 phen. No attempt was made to separate the employed for the preparation of the complexes OC-6-23-[Rh(2-S- geometricalisomersofcomplexescontainingtheseligands. py) (pp)]Cl 6 (pp=dpq) and 7 (pp=dppz) using pyridine-2- 2 thiolate as the second bidentate ligand. All the compounds were 2.2.DNAbindingstudies characterizedby1HNMRandpositive-ionLSIMS(Liquidsecondary ionmassspectrometry)andgavesatisfactorymicroanalyses. Measurement of the thermal denaturation temperature T m for Structuresofthecomplexes5and7basedonDFTcalculationsare calf thymus DNA (CT DNA) in the presence of a transition metal depicted in Figs. 1 and 2. On comparison of their geometries, it is complexprovidesastraightforwardmeansofgaginganystabiliza- apparentthattheorientationoftheH6protonofoneofthe[2-S-py]− tion of the double helix due to a specific binding interaction. No ligandsabovethebipyridinepartofthedppzligandincomplex7may changewasdetectedfortheT m value(67°Cina10mMphosphate restrictitspossibleintercalationintotheDNAduplex.Inaccordance bufferatpH=7.2)ofCTDNAinthepresenceofcomplexes1and2 withanOC-6-23coordinationgeometry(StranstoN(py)andS′trans to N(pp)), two separate 1H NMR resonances with equal integral valueswererecordedfortheH2/9(2d),H4/7(2d)andH3/8(2dd) Scheme2.Structuresofcomplexes6and7. Fig.2.DFTstructureoftheOC-6-23isomerofcomplex7. F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 993 Fig.3.1HNMRspectrumoftheOC-6-23isomerofcomplex6ind6-DMSO. Fig.4.CDspectraofCTDNAaloneandwiththecomplexes3,4and5atr=0.2ina 10mM phosphate buffer (pH=7.2) after an incubation period of 60min. Molar ellipticities[Θ]aregivenintheunitsdeg.cm2dmol−1×103. andonlyamodestincreaseof1°Cinthepresenceof3,allofthese measurements being performed at a [complex]/[DNA] ratio of fromastericallyinducedchangeintheorientationofthedppzligand r=0.2where[DNA]=M(basepairs).Inaccordancewithapossible relativetothebasestack[21]incomparisontothedpqcomplex4.No stabilizing intercalative binding mode, larger increases of respec- significant change in the CT DNA spectrum was observed in the tively4and2°Cwerecausedbythedpqanddppzcomplexes4and presenceofthepyridine-2-thiolatocomplexes6and7,whichsupports 5 at the same ratio. In striking contrast to these dimethyldithio- theUV/visiblefindingsthattheirdpqanddppzligandsdonotexhibit carbamatocomplexes,ΔT valuesofzerowererecordedforCTDNA significantintercalation. m in the presence of the analogous pyridine-2-thiolato complexes 6 CorroborativeevidenceforDNAintercalationcanoftenbeobtained (pp=dpq)and7(pp=dppz). from viscosity measurements [22,23]. Insertion of large aromatic Transition metal dppz complexes exhibit characteristic absorption ligands such as dpq and dppz between adjacent nucleobase pairs bands at about 364 and 383nm, which are due to π–π* transitions leadstoalengtheningandstiffeningofthedoublehelixandthese centeredinthephenazinepartofthepolypyridylligand[13].Pronounced changesarereflectedinanincreaseinDNAviscosity.Fig.5illustrates decreasesinabsorptionandbathochromicshiftsforthesemaximaare the dependence of the reduced viscosity function (η/η )1/3 on the 0 generallyindicativeofdppzintercalationbetweenthebasepairsofthe [complex]/DNAratiorforcomplexes4and5.As(η/η )1/3isequalto 0 DNA double helix [14,15]. The ΔA/A values of respectively −19 and (L/L ),whereListhelengthofintercalatedDNAandL isthelengthof 0 0 −22%andbathochromicshiftsof+2nm,whichwererecordedforthe DNAalone[22],aslopeof1.0istobeexpectedforidealintercalation absorptionbandsofcomplex5ata[complex]/[DNA]ratioof0.2,are, atlowerrvalues.Thisvaluereflectstheclassicalmodelofintercalation, therefore,indicativeofpossibleintercalativebinding.Ontheotherhand, wherethehelixislengthenedby3.4Åperintercalatedaromaticmoiety. themarginalΔA/Achangesofonly−3and−4%forthesemaximainthe Infact,amuchwiderrangeofpossiblehelixextensionsfrom2.0to analogous complex 7/CT DNA mixture suggest that the pyridine-2- 3.7Åhasbeenindicatedbyelectricdichroismmeasurements[24].A thiolatocomplexdoesnotintercalatetosignificantextent. slope of 0.93 was determined for ethidium bromide as a standard AnegativeCDbandat246nmduetothehelixBconformationand strongintercalatorusingourexperimentalsetup.Thevalueof0.92 a positive band at 275nm caused by nucleobase pair stacking forCTDNAinthepresenceofdpqcomplex4(Fig.5)is,therefore,in representthecharacteristicfeaturesinthecirculardichroism(CD) spectrum of CT DNA [16]. Addition of 1 or 2 to CT DNA has no significant effect on this typical spectrum. Fig. 4 depicts the CD spectra of the duplex in the presence of complexes 3, 4 and 5 at r=0.2.Theformertwocompoundscausedsignificantenhancements inthemolarellipticity[Θ]ofthebase-stackingsignalatabout275nm which were accompanied by shifts to lower wavelengths. Similar changes have often been recorded for intercalative binding by transitionmetalcomplexeswithextendedpolypyridylligands(e.g. pp=dpq,dppz),ashavenegativeinducedCDbands(ICDs)atabout 295nm,suchasthoseinvokedby3and4[17,18].However,these ICDs could also be caused by the polypyridyl ligands adopting a specificorientationrelativetotheduplexinagroovebindingmode.It isinterestingtonotethatMahadevanandPalaniandavar[19]have reportedtheinductionofanegativeCDbandatabout290nmfora mixture of [Cu(5,6-Me phen) ]2+ and CT DNA, which they inter- 2 2 pretedasindicatingapartialconformationalchangetoZDNA,whose characteristicCDbandsareat240(minimum),265(maximum)and 292nm(minimum)[20].Incontrastto3and4,thedppzcomplex5 causedonlyanarrowingofthepositiveCDbandatabout275nmbut nonegativeICDatabout295nm.Thisissurprising,inviewofthe Fig.5.Dependenceof(η/η 0)1/3onr{r=[complex]/[DNA]where[DNA]isgiveninM (basepairs)}forviscositymeasurementsofCTDNAwithcomplexes4,5and7ina significantdecreasesinΔA/Aobservedforitsabsorptionmaximaat 10mMphosphatebuffer(pH=7.2).Regressionlinesaredepictedtounderlinethe 364and383nmonmixingwithCTDNA,butcouldpossiblyresult lineardependence. 994 F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 accordance with this binding mode, as was also indicated by the pronouncednegativeICDat295nmandtheT increaseof+4°C.A m significantlylowervalueof0.63wasobtainedforCTDNAwithdppz complex 5, possibly due to steric contacts preventing optimum intercalativeoverlapforthelargerpolypyridylligand.Itisinteresting to note that a similar slope value of 0.69 was determined for CT DNAwiththe5,6-Me phencomplex3.Thissuggeststhatitssmaller 2 phenathroline ligand may possibly also participate in intercalative binding.However,theobservedDNAlengtheningcouldalsoresult fromspecificgroovebindingand/or,inpart,fromaB→Zconforma- tionalchange.Muchlowerslopesof0.45and0.20wereobtainedfor thepyridine-2-thiolatocomplexes6and7.Takentogetherwiththe CDandUV/visiblespectraformixturesofthesecomplexeswithCT DNA,suchsmallerviscosityincreasessuggestthatDNAintercalation maynotbeofsignificanceforthepyridine-2-thiolatocomplexes.It seemslikelythatthepresenceofanH6protonabovethebipyridine partofthedpqanddppzligandsintheseOC-6-23isomers(Fig.2) will restrict the extent of any possible stacking overlap with the DNAbasepairs.IfDNAisindeedasignificantintracellulartargetfor Fig.6.Standardcellimpedance(%)forMCF-7cellstreatedwith1–10μMconcentrationsof compound6overa36hperiod(5–41h).RM=runningmedium. suchtris(chelate)complexes,thenthemetallointercalators4and5 mightbeexpectedtoexhibithigherantiproliferativeactivitythan6 and7. intercalationmaynotbeofsignificanceforthemodeofantiprolifera- tiveactionofthetris(chelate)complexes. InourrecentstudiesoncytostaticRu(II)andRh(III)polypyridyl 2.3.Antiproliferativeandcellularuptakeproperties complexes[9,12,18]wehaveobservedadirectrelationshipbetween the level of cellular uptake and their antiproliferative activity. To Invitrostudiesontheantiproliferativepropertiesofcomplexes3– ascertain the importance of this factor for the kinetically inert tris 7wereperformedusingcrystalvioletassaysfortheadhesivehuman (chelate) complexes, we employed atomic absorption spectroscopy cancer cell lines MCF-7 (breast adenocarcinoma) and HT-29 (colon (AAS)tomeasuretherhodiumcontentofHT-29cellsexposedto1μM adenocarcinoma). The resulting IC values are listed in Table 1 50 ofthecomplexes4,6(pp=dpq)and7(pp=dppz)overaperiodof togetherwiththeliteraturevaluesforthe5,6-Me 2 phen[25],dpqand 6h.Initialpilotexperimentswith0.1μMsolutionshadaffordedonly dppz ligands [9]. All five complexes are highly active towards the lowuptakevalues.Followingincubationwiththecomplexes,thecells cancercelllinesandexhibitIC 50 valuesintherange0.055–0.285μM werewashedwithphosphatebufferedsalinetoremovepossiblecell thataresome102–103timeslowerthanthosepreviouslyrecordedfor fragments,harvestedbytrypsinizationandlysedbyultrasonication. substitutionally inert tris(chelate) complexes! For instance the IC 50 TherhodiumcontentoftheresultingsampleswasdeterminedbyAAS valuesof0.104and0.285μMforcomplex4towardsMCF-7andHT-29 and the protein content by the Bradford method. The latter values cellsaresome498-and218-foldlowerthanthosereportedfor[Ru indicatedthatcelldeathduringthe6hincubationperiodwasnotof (bpy) 2 (dpq)]2+towardsthesamecelllines[5].Theanalogousactivity significance. In accordance with the previous studies, the cellular ratiosforcomplex5versus[Ru(bpy) 2 (dppz)]2+are1640and465.In uptake(expressedasngrhodiumpermgcellprotein)wasinfluenced analogytothekineticallyinertrhodium(III)complexes[(η5-C Me ) 5 5 bythesizeoftherespectivepolypyridylligandwithcloselysimilar Rh{(Me N) CS}(pp)] (CF SO ) (pp=dpq, IC =5.3 and 10.7μM; 2 2 3 3 2 50 valuesof59.9(3.8)and60.1(3.0)ng/mgbeingrecordedfor4and6, pp=dppz,IC =1.5and4.3μM)[9],theantiproliferativeactivityof 50 respectively,incomparisontoahigheruptakeof80.1(5.2)ng/mgfor thedimethyldithiocarbamatocomplexes4and5isclearlycorrelated 7. It is interesting to note that these values are very close to those to the size of the respective polypyridyl ligand (dpqbdppz). In obtainedfortheneutralrhodium(III)complexesmer-[RhCl (DMSO) contrast,thepyridine-2-thiolatocomplex6(pp=dpq)issignificantly 3 (pp)](pp=dpq,dppz),whichalsoexhibitastrongantiproliferative more active towards MCF-7 cells and only marginally less effective activity towards HT-29 cells. The respective cell uptake and IC 50 towardsHT-29cellsthanitsdppzcounterpart.Itisinterestingtonote values were 53.6(2.0) ng/mg and 0.069(0.021) μM for mer-[RhCl 3 thattheMCF-7andHT-29IC 50 valuesforthenon-intercalator6are (DMSO)(dpq)] and 69.8(1.8) ng/mg and 0.073(0.017) μM for mer- 2.2and3.7timeslowerthanitsdimethyldithiocarbamatocounterpart [RhCl (DMSO)(dppz)][12].Theobservedcellularlevelsfor4,6and7 3 4,which,incontrast,isametallointercalator.ThissuggeststhatDNA can be classified as being very high when compared to established metallodrugs.Ontakingindividualcellularparametersintoaccount,it canbeestimatedthat1.0ngrhodiumpermgcellproteincorresponds to 1.9μM cellular rhodium in HT-29 cells [12], which means that Table1 compound7,forinstance,isaccumulated152foldinthecancercells. IC50valueswithestimatedstandarddeviationsinparenthesesforthecomplexes3–7 towardsMCF-7andHT-29cells. In comparison, the platinum drugs cisplatin, carboplatin and oxaliplatinareonlytakenupata1.5–6foldlevel[26]. IC50[μM] The much higher IC values of 62.1(3.9) and 26.9(0.4) μM for 50 Compound pp MCF-7 HT-29 thetris(chelate)complexes[Ru(bpy) (pp)]Cl (pp=dpq,dppz)towards 2 2 3 5,6-Me2phen 0.128(0.064) 0.188(0.030) HT-29cellscanbeexplained,atleastinpart,bytheirmuchdiminished 4 dpq 0.104(0.015) 0.285(0.017) levelsofcellularuptake,whicharesome285and203timeslowerthan 5 dppz 0.055(0.005) 0.058(0.008) for 6 and 7. Although the lower overall charge of the monocationic 6 dpq 0.059(0.007) 0.078(0.014) rhodium(III)tris(chelate)complexeswouldbeexpectedtosignificantly 7 dppz 0.109(0.012) 0.055(0.036) 5,6-Me2phen[25] 0.34(0.08) 0.28(0.16) improve their cellular uptake and therefore their antiproliferative dpq[9] 6.7(2.0) 7.0(2.2) activityrelativetothedicationic ruthenium(II) complexes[Ru(bpy) 2 dppz[9] 0.8(0.6) 1.8(0.2) (pp)]Cl ,thepossibleinfluenceofthecentralmetalatomswiththeir 2 cisplatin[9] 2.0(0.3) 7.0(2.0) differingoxidationstatesshouldnotbeneglected. F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 995 2.4.Cellularmetabolism The metabolism of MCF-7 cells in the presence of the highly cytostaticcomplex6wasmonitoredonlinewithaBionasbiosensor chip system, whose silicon chip contains interdigitated electrode structuresformeasuringthecellularimpedance[27],miniatureClark- typeoxygenelectrodesforfollowingthecellularoxygenconsumption [28]andion-sensitivefieldeffecttransistorstorecordextracellularpH changes[29].IncontrasttothecrystalvioletassaywheretheMCF-7 cellsareplatedatamediumcelldensityandduplicatewithinabout 30h,cellsareplatedathighdensityintheBionassystem.Asaresult theirrateofproliferationismuchlower.Therecordedcellimpedance doesnotreflectthecellnumberbutrathertheintegrityofthecell layer,i.e.itprovidesameasureofcelldeathratherthanproliferation inhibition. Whereas concentrations close to the IC value should 50 affect the modest proliferation rate in the Bionas system, higher concentrationsaregenerallynecessarytoinvokemeasurableresponses duetoapoptoticcelldeathornecrosis.AsillustratedinFig.7,significant changesinMCF-7cellimpedanceandrespirationwereobservedfora 1μMsolutionofcomplex6withinthe36htreatmentperiod.Higher2.5, Fig.8.Standardextracellularacidificationrates(%)forMCF-7cellstreatedwith1–10μM 5and10μMsolutionswerestudiedforcomparisonpurposes. concentrationsofcompound6overa36hperiod(5–41h).RM=runningmedium. Followingexposureperiodsof6and8htorespectively10and 5μMsolutionsofcomplex6,theimpedanceofMCF-7celllayersfell rapidly to a steady percentage level of about only 10–15% of the solutionsofcomplex6(Fig.8),whichsuggestthatthecellsenhanced initiallevel(Fig.6).Thesefinallevelswerereachedafterrespectively glycolysis to compensate for the reduced respiration. After longer 11and15hoftreatment.Longerinductionperiodsandslowerrates exposureperiods,concentration-dependentdecreasesintheacidifi- of impedance decline were recorded for the weaker 2.5 and 1μM cation rate became apparent, that can be ascribed to cell dying or solutions.Thedramaticlossofimpedanceforthemoreconcentrated death.Wehavepreviouslyshownthattherespirationinhibitioncaused solutions indicates morphological changes and/or changes in the bygoldcomplexesinMCF-7cellsalsoleadstoasignificantincreasein cellularadhesionproperties,suchascell–cellandcell–matrixcontacts, glycolysis[30]. correspondingtotheinductionofcelldeath.Itisinterestingtonotethat AstheMCF-7cellmetabolismstudiesindicatedthattheimmedi- ananalogoussensor-chipexperimentforHT-29cellsexposedtoamuch ateanddramaticreductionincellularoxygenconsumptioncaused moreconcentrated100μMsolutionof[Ru(bpy) (dppn)]Cl registered by exposure to complex 6 leads to the induction of cell death, we 2 2 an immediate linear but slower decrease in impedance [5], which evaluatedtheeffectof6ontherespirationofisolatedmice(wildtype, dropped to about30%of itsoriginal levelafter24h oftreatment. In C57BL/6)livermitochondria.Inthisassay,respirationoffunctionally contrast,effectivelynochangeinHT-29cellimpedancewasrecorded activemitochondrialeadstoaloweringintheoxygenlevelofthebuffer for100μMsolutionsoftheanalogousdpqanddppzcomplexes[5].The medium,thatcanbeantagonizedbyimpairmentofmitochondriavital authors suggested that their observations could indicate a direct functions.AsillustratedinFig.9,complex6causesadose-dependent interactionofthedppncomplexwithconstituentsofthecellmembrane inhibition of mitochondrial respiration, which indicates that a mito- orembeddedmembraneproteins. chondrialpathwaymaybeinvolvedinthecell-deathprocess. Treatmentwithcomplex6ledtoanimmediateandpronounced decreaseinoxygenconsumptionfortheMCF-7cells,whichwasdose- 2.5.ApoptosisinductioninJurkatcells dependentandcorrelatedtothelossofcell-layerimpedancedepicted inFig.6.Initialincreasesintheextracellularacidificationratewere Itwasalsoofinteresttoestablishwhetherthemonocationictris observedfortheMCF-7cellsontreatmentwith1.0,2.5and5.0μM (chelate)complexesexhibitastrongantiproliferativeactivitytowards malignant non-adhesive cells as has recently been reported for the rhodium(III)complexes[RhCl (DMSO)(pp)](pp=5,6-Me phen,dpq) 3 2 [25,31],[(η5-C Me )RhCl(5,6-Me phen)](CF SO )[32]and[RhCl([9] 5 5 2 3 3 aneS )(bpm)]Cl (bpm=2,2′-bipyrimidine) [33]. Staining with 3 2 Annexin V-FITC, then counterstaining with propidium iodide (PI), providesa convenientmethodforestimatingthenumberofviable, apoptoticandnecroticcellsfollowingtreatmentwithacytotoxicagent. AnnexinV-FITCbindsspecificallytothephospholipidphosphatidylser- inethatisexposedtotheouterleafletoftheplasmamembraneduring apoptosis[34,35].TheDNAintercalatorPI,ontheotherhand,canonly penetratedeadordyingcellsthathavelosttheirmembraneintegrity.PI positivityis, therefore,anindicator foreithercell necrosis or for the presenceoflateapoptoticcells,whoseplasmamembraneisnolonger intact[36]. Fig. 10 illustrates the cell-state distribution for Jurkat leukemia cellsafter48htreatmentswiththedpqcomplexes4and6atselected concentrations. These were chosen to span the range from values belowthosethatcause50%inhibitionofproliferationuptoabout10 timesthisvalue.Althoughbothcompoundsexhibitedapronounced Fig.7.Standardrespirationrates(%)forMCF-7cellstreatedwith1–10μMconcentrations antiproliferativeeffect,theactivityofthepyridine-2-thiolatocomplex ofcompound6overa36hperiod(5–41h).RM=runningmedium. 6 is clearly much more pronounced, which is also the case for its 996 F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 Fig.9.Respirationoffreshlyisolatedmouselivermitochondria.Mitochondrialactivityleadstoadecreaseintheoxygensaturationoftherespirationbuffer.Compound6causesa concentration-dependentinhibitionofmitochondrialrespirationincomparisontothenontreatedcontrol.Furthercontrols:rotenone(aninhibitoroftherespiratorychaincomplex I),CCCP(carbonylcyanide3-chlorophenylhydrazone,anuncouplerofoxidativephosphorylation),DMF(correspondingtothesolventconcentrationinthe10μMsolutionof6). apoptosis-inducingproperties.Whereas87%oftheJurkatcellswere 3.Conclusions still viable following a 48h treatment with a 2.5μM solution of complex4,thispercentagefelltoonly5%afterincubationwiththe Thesubstitutionallyinerttris(chelate)complexes[Rh{(CH ) NCS } 3 2 2 2 sameconcentrationofcomplex6.88%ofthecellsexhibitedapoptosis (pp)]Cl3–5(pp=5,6-Me phen,dpq,dppz)and[Rh(2-S-py) (pp)]Cl6 2 2 inthiscase,comparedtoonly7%thatwereinanecroticstate. and 7 (pp=dpq, dppz) exhibit a pronounced antiproliferative effect Reactive oxygen species (ROS) are by-products of the aerobic towardstheadherenthumancancercelllinesMCF-7andHT-29asdoes metabolismtakingplaceinthemitochondria.Only1–3%ofhealthycells the selected complex 6 towards non-adherent Jurkat leukemia cells. typicallyexhibithighROSlevels[37].HigherROSconcentrationshavea Uptake studies revealed that their high general levels of activity destructiveeffectonDNAandmanyproteinsandtheresultingoxidative correlatedirectlywithhighintracellularrhodiumlevels.Whereasthe stresscantriggerapoptosis[38,39].LevelsofROScausedbycomplexes4 IC valuesforHT-29cellsincubatedwiththedppzcomplexes5and7 50 and 6 in Jurkat cells (Fig. 11) were estimated by fluorescence are closely similar (0.058 and 0.055μM), the pyridine-2-thiolato measurements at 564–606nm following treatment of the incubated complex 6 (pp=dpq, IC =0.078μM) is considerably more active 50 cellsuspensionswithdihydroethidiuminthedarkfor15min.Thesame thanitsdimethyldithiocarbamatoanalog4(pp=dpq,IC =0.285μM). 50 concentrationsofcompounds4and6wereemployedasfortheAnnexin AsUV/visibleandCDstudiesclearlyindicatethatcomplexes6and7,in V-FITCassaysofFig.10.ThehighROSlevelsareclearlydose-dependent contrastto4and5,arenotmetallointercalators,itcanbeassumedthat for both compounds and significantly higher for complex 6 at all DNAintercalationbythecompounds' polypyridyl ligandswill notof concentrations.AsignificantincreaseinthehighROSlevelsisapparent centralimportancefortheirmodeofcytotoxicaction. evenatlowconcentrationsof4(0.62μM)and6(0.16μM)forwhich Compound 6 invokes extensive apoptosis in Jurkat cells, which antiproliferativeeffectsbutonlynegligibleapoptosisornecrosiswere correlateswiththeobserveddose-dependentconcentrationsofreactive observed(Fig.10).AcorrelationbetweenthepercentageofJurkatcells oxygen species in these leukemia cells. This suggests that oxidative withhighROSlevelsandthenumberofcellsundergoingapoptosisis stressmay playa central role in initiatingthecell death program. It apparentoncomparisonofFigs.10and11.Itisinterestingtonotethat shouldbenotedthatlightirradiationisnotrequiredtostimulatethe increased formation of ROS has recently been implicated in the production of high ROS concentrations in the presence of the tris mechanismofantiproliferativeactionofaquaternarydppzderivative, (chelate)complexes4and6.Theobservationofanimmediatedose- albeitonlyuponlightirradiation[40]. Fig.10.StateofJurkatcellsonthebasisofanAnnexinV/PIbindingassayfollowing treatment with complexes 4 and 6 for 48h. Cells were also treated with 5μM Fig. 11. Levels of intracellular reactive oxygen species (ROS) in Jurkat cells after camptothecin(CMPT)andDMSO(mock)ascontrols. incubationwithcomplexes4and6for48h. F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 997 dependent reduction of oxygen consumption by MCF-7 cells and 12H,carbamateCH ),8.08(dd,2H,H3/8),8,85(d,2H,H4/7),9.60(d,2H, 3 isolatedmicemitochondriaontreatmentwithcomplex6indicatesthat H2/9). the induced apoptosis is mediated via the intrinsic mitochondrial pathway. 4.1.4.[Rh{(CH ) NCS } (dpq)]Cl(4) 3 2 2 2 Preparation as for 1 with [RhCl (DMSO)(dpq)] (100mg, 3 4.Experimental 0.192mmol) and sodium dimethyldithiocarbamate (55.0mg, 0.384mmol).Yield:58%(67.8mg).C H ClN RhS ;M=611.0g/mol; 20 20 6 4 Solventsweredriedanddistilledbeforeuse.1HNMRspectrawere Anal.Found(%):C39.5;H3.3;N13.6;Calc.(%):C39.3;H3.3;N13.7. recordedonaBrukerDRX200spectrometerwiththeresidual1Hsignal LSIMS: m/z(%) 575(100) [M-Cl]+, 455(35) [M-HCl-(CH ) NCS ]+. 1H 3 2 2 ofthedeuteratedsolventbeingusedforchemicalshiftcalculation.The NMR(CD Cl ,200MHz,30°C):δ=3.15,3.42(2s,12H,carbamateCH ), 2 2 3 splittingsoftheprotonresonancesaredefinedass=singlet,d=doublet 8.13(dd,2H,H3/8),8.27(s,2H,H11/12),8.86(d,2H,H4/7),9.70(d,2H, andm=multiplet.LSIMS(LiquidSecondaryIonMassSpectrometry) H2/9). datawereregisteredwithaFisonsVGAutospecinstrumentemployinga cesiumiongun(voltage17kV)and3-nitrobenzylalcoholastheliquid 4.1.5.[Rh{(CH ) NCS } (dppz)]Cl(5) 3 2 2 2 matrix. An Analytikjena SPECORD 200 was employed for UV/visible Preparation as for 1 with [RhCl (DMSO)(dppz)] (100mg, 3 measurements and CD spectra were recorded on a Jasco J-715 0.176mmol) and sodium dimethyldithiocarbamate (50.4mg, instrument. Elemental analyses were performed with a Vario El 0.352mmol).Yield58%(64.1mg).C H ClN RhS M=661.1g/mol; 24 22 6 4; (ElementarAnalysensysteme).RhCl ·3H OwaspurchasedfromChem- Anal.Found(%):C42.0;H3.7;N13.0;Calc.(%):C42.3;H3.8;N12.4. 3 2 pur and 2,2′-bipyridine (bpy), 1,10-phenanthroline (phen) and 5,6- LSIMS: m/z(%) 625(100) [M-Cl]+, 505(55) [M-HCl-(CH ) NCS ]+. 1H 3 2 2 dimethyl-1,10-phenanthroline (5,6-Me phen) from Acros. Sodium NMR(CD Cl ,200MHz,30°C):δ=3.16,3.44(2s,12H,carbamateCH ), 2 2 2 3 dimethyldithiocarbamate and 2-mercaptopyridine (H[2-S-py]) were 8.12(dd,2H,H3/8),8.21(d,2H,H11/14),8.84(d,2H,H12/13),9.74(d, obtained from Aldrich. Solvents were purchased from J.T. Baker and 2H,H4/7),9.95(d,2H,H2/9). DEUTEROGmbH.Thepolypyridylligandsdpq[41]anddppz[42]were preparedinaccordancewithliteratureprocedures,aswerethestarting 4.1.6.OC-6-23-[Rh(2-S-py) (dpq)]Cl·2HCl·3H O(6) 2 2 complexes[RhCl (DMSO)(pp)]withpp=bpy,phen,5,6-Me phen,dpq [RhCl (DMSO)(dpq)](100mg,0.192mmol)andtwoequivalents 3 2 3 and dppz [12]. DFT calculations were performed with Gaussian 03, of 2-mercaptopyridine (42.2mg, 0.384mmol) were dissolved in RevisionE.01[43].Geometrieswerecalculatedusingthehybriddensity 15ml methanol and refluxed for 8h. The solvent was removed in functionalB3LYP,theStuttgartRSC1997ECPbasissetforrhodiumand vacuumtoaffordayellowsolidthatwaswashedwithmethanoland 6–31G(d)forallotheratoms.TheStuttgartRSC1997ECPbasissetwas dried in vacuum. Yield: 80% (108.6mg). C H Cl N RhS ·3H O; 24 18 3 6 2 2 obtainedviatheEMSLwebsiteBasisSetExchange[44,45].Converged M=717.9g/mol;Anal.Found(%):C40.0;H3.0;N11.4;Calc.(%):C geometrieswerecharacterizedasminimabytheabsenceofimaginary 40.1;H3.4;N11.7.LSIMS:m/z(%)556(47)[M-Cl]+,445(100)[M- modes in the calculated vibrational frequencies. The graphical user HCl-(2-S-py)]+.1H NMR(DMSO-d ,200MHz,30°C):δ=6.78(m, 6 interfaceGabedit[46]wasusedforoutputandstructurehandling. 2H,2-S-py),7.29(m,2H,2-S-py),7.42(m,2H,2-S-py),7.67(m,2H, 2-S-py),8.43,8.53(2dd,2H,H3/8),9.39(s,2H,H11/12),9.75,9.85(2d, 4.1.Preparationof1–7 2H,H4/7),10.12,10.22(2d,2H,H2/9). 4.1.1.[Rh{(CH ) NCS } (bpy)]Cl(1) 4.1.7.OC-6-23-[Rh(2-S-py) (dppz)]Cl·2HCl·3H O(7) 3 2 2 2 2 2 [RhCl (DMSO)(bpy)](100mg,0.225mmol)andtwoequivalents Preparation as for 6 with [RhCl (DMSO)(dppz)] (100mg, 3 3 of sodium dimethyldithiocarbamate (64.4mg, 0.45mmol) were 0.176mmol)and2-mercaptopyridine(40.0mg,0.352mmol).Yield: dissolvedin15mlofmethanolandrefluxedfor6h.Aftercoolingto 74% (102.8mg). C H Cl N RhS.3H O; M=767.9g/mol; Anal. 28 20 3 6 2 2 room temperature, the resulting precipitate was dried in vacuum. Found (%): C 43.4; H 3.0; N 10.3; Calc. (%): C 43.8; H 3.4; N 10.9. Following washing with ice-cold water, it was redissolved in LSIMS: m/z (%) 606(58) [M-Cl]+, 495(100) [M-HCl-[(2-S-py)]+. 1H methanolanddriedagaininvacuum.Yield:60%(72.1mg).C H NMR(DMSO-d ,200MHz,30°C):δ=6.77(m,2H,2-S-py),7.30(m, 16 20- 6 ClN RhS ; M=535.0g/mol; Anal. Found (%) C 36.3; H 4.1; N 10.3; 2H,2-S-py),7.42(m,2H,2-S-py),7.66(m,2H,2-S-py),8.22(2dd,2H, 4 4 Calc.(%):C35.9;H3.8;N10.5.LSIMS:m/z(%)499(100)[M-Cl]+,378 H12/13),8.42(2dd,2H,H3/8),8.55(2d,2H,H11/14),9.85,9.94(2d, (43) [M-HCl-(CH ) NCS ]+. 1H NMR (CD Cl , 200MHz, 30°C): 2H,H4/7),10.12,10.22(2d,2H,H2/9). 3 2 2 2 2 δ=3.17,3.37(2s,12H,carbamateCH ),7.72(dd,2H,H3/8),8.23(d, 3 2H,H5/6),8.73(dd,2H,H4/7),9.38(d,2H,H2/9). 4.2.DNAbindingstudies 4.1.2.[Rh{(CH ) NCS } (phen)]Cl(2) ThethermaldenaturationtemperaturesT of1:5complex:DNA 3 2 2 2 m Preparation as for 1 with [RhCl (DMSO)(phen)] (100mg, mixtures [DNA concentration=M(base pairs)] were determined 3 0.214mmol) and sodium dimethyldithiocarbamate (61.3mg, for 1–7 in a 10mM phosphate buffer at pH=7.2. Melting curves 0.428mmol).Yield:46%(54.8mg).C H ClN RhS ;M=559.0g/mol; wererecordedat1°Cstepsforthewavelengthλ=260nmwithan 18 20 4 4 Anal.Found(%):C38.2;H3.7;N10.2;Calc.(%):C38.6;H3.6;N10.0. Analytikjena SPECORD 200 spectrometer equipped with a Peltier LSIMS: m/z(%) 523(100) [M-Cl]+, 403(43) [M-HCl-(CH ) NCS ]+. 1H temperature controller. T values were calculated by determining 3 2 2 m NMR(CD Cl ,200MHz,30°C):δ=3.16,3.42(2s,12H,carbamateCH ), the midpoints of melting curves from the first order derivatives. 2 2 3 8.07(dd,2H,H3/8),8.22(s,2H,H5/6),8.75(d,2H,H4/7),9.63(d,2H, TheexperimentalΔT valuesareestimatedtobeaccuratetowithin m H2/9). +/−1°C. Concentrations of CT DNA were determined spectrophoto- metricallyusingthemolarextinctioncoefficientε =13200M−1cm−1 260 4.1.3.[Rh{(CH ) NCS } (5,6-Me phen)]Cl(3) [47]. 3 2 2 2 2 Preparationasfor1with[RhCl (DMSO)(5,6-Me phen)](100mg, Viscosities for mixtures of complexes 3–7 with sonicated DNA 3 2 0.202mmol) and sodium dimethyldithiocarbamate [57.8mg, were determined using a Cannon-Ubbelhode semi-micro dilution 0.404mmol).Yield:49%(57.8mg).C H ClN RhS ;M=587.1g/mol; viscometer(SeriesNo75,CannonInstrumentsCo)heldataconstant 20 24 4 4 Anal.Found(%):C41.3;H4.5;N9.3;Calc.(%):C40.8;H4.5;N9.5. temperatureof25°Cinawaterbath.Theviscometerinitiallycontained LSIMS: m/z(%) 551(100) [M-Cl]+, 431(46) [M-HCl-(CH ) NCS ]+. 1H 2mlof0.4mMsonicatedDNAsolutionina10mMphosphatebuffer 3 2 2 NMR(CD Cl ,200MHz,30°C):δ=2.87(s,6H,CH 5/6),3.15,3.42(2s, (pH=7.2). 0.2mM complex solutions also containing 0.4mM 2 2 3 998 F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 sonicatedDNAwereaddedinincrementsof100μlfromamicropipette. incubatorat37°C/5%CO and95%humidityfor24huntil80–90% 2 Solutions were passed throughfilterstoremove particulate material confluency was reached. Sensor chips with cells were then priortouse.Reducedviscositiesηwerecalculatedbyliteraturemethods transferred to the Bionas 2500 analyzer in which medium was [22]andplottedas(η/η )1/3(η =reducedviscosityoftheDNAsolution continuously exchanged in 10min cycles (3min exchange of 0 0 intheabsenceofcomplex)againstrforrod-likeDNA(approximately medium and 7min without flow) during which the parameters 600basepairs). weremeasured.Therunningmedium(RM)duringtheanalysiswas DMEMwithoutcarbonatedbuffer(PANCat.Nr.P03-0010)andonly 4.3.Antiproliferativeactivity weakly buffered with 1mM Hepes, reduced FCS (0.1%) and low glucose (1g/l). For drug activity testing, the following steps were MCF-7 breast cancer and HT-29 colon cancer cells were main- followed: (a) 5h equilibration with running medium with 3 and tained in DMEM High Glucose (PAA) supplemented with 50mg/l 7minflow/stopincubationintervals,(b)36hdrugincubationwith gentamycinand10%(v/v)fetalcalfserum(FCS)at37°C/5%CO and substancesfreshlyaddedinmediumatindicatedconcentrationsalso 2 passaged once a week according to standard procedures. The with the same flow/stop cycles. At the end of each experiment, cell antiproliferative effects of compounds 3–7 were determined by an layerswereremovedbyadditionof0.2%TritonX-100toobtainabasic established procedure using crystal violet assays for treated MCF-7 signalwithoutlivingcellsonthesensorsurfaceasanegativecontrol. and HT-29 cells [48]. Cells were suspended in cell culture medium (MCF-7: 10000 cells/ml; HT-29: 3000 cells/ml) and 100μl aliquots 4.6.AnnexinV-propidiumiodidebindingassay thereofwereplatedin96well-platesandincubatedat37°C/5%CO 2 for72h(MCF-7)or48h(HT-29).Stocksolutionsofthecompounds Jurkat cells were purchased from the German Collection of 3–7 in DMF were freshly prepared and diluted with cell culture Microorganisms and Cell Cultures (DSMZ, Braunschweig) and main- mediumtothedesiredconcentrations(finalDMFconcentration:0.1% tainedat37°CinRoswellParkMemorialInstitute(RPMI)1640medium (v/v)). The medium in the plates was replaced with medium with10%heat-inactivatedFCS(bothfromPAA).Uptoamaximumof containing the compounds in graded concentrations (six replicates, 4×105cellsperwellwereincubatedwiththesubstances4and6at 200μlperwell).Afterfurtherincubationfor96h(MCF-7)or72h(HT- differentconcentrationsfor48h.300μlaliquotsofcellsuspensionwere 29),thecellbiomasswasdeterminedbycrystalvioletstainingand takenfromeachwellandthecellswereresuspendedin50μlAnnexinV theIC valueswerecalculatedasthoseconcentrationscausing50% bindingbuffer.Then2.5μlAnnexinV-FITC(eBioscience,SanDiego,CA) 50 inhibition of cell proliferation. Results were obtained as average wereaddedandthesampleswerevortexedandincubatedfor10minat valuesfrom2to3independentexperiments. r.t.Followingadditionof450μlofbindingbuffer,thesuspensionswere transferredtoFACStubes,mixedwith1.5μlpropidiumiodide(1mM) 4.4.Cellularuptakestudies andfurtherincubatedfor5–10minatr.t.FACSanalyseswereperformed withtheFACS®Calibur(BectonDickinson,Heidelberg,Germany)and ForcellularuptakestudiesHT-29cellsweregrownuntilatleast70% theCellQuest™Proanalysissoftwareusinganexcitationwavelengthof confluencyin175cm2cellcultureflasks.Stocksolutionsofcomplexes4, 488nmandemissionsettingsof515–545nm(FL1channel)forAnnexin 6 and 7 in DMF were freshly prepared and diluted with cell culture V-FITCdetectionand564–606nm(FL2channel)forpropidiumiodide mediumtothedesiredconcentrations(finalDMSOconcentration:0.1% detection. v/v,finalcomplexconcentration:1.0μM).Thecellculturemediumof the cell culture flasks was replaced with 10ml of the cell culture 4.7.MeasurementofintracellularROS mediumsolutionscontaining4,6or7andtheflaskswereincubatedfor 6hat37°C/5%CO .Afterwardsthemediumwasremovedandthecells Jurkat cells were incubated with different concentrations of 2 werewashedwithphosphatebufferedsalinepH7.4.Aftertrypsiniza- compounds 4 and 6. The cells were collected after 48h treatment, tion,cellpelletswereisolatedbycentrifugation,resuspendedin1–5ml centrifugedat0.2×g(1500rpm)andresuspendedinFACSbuffer(D- twicedistilledwater,lysedbyultrasonicationandappropriatelydiluted PBS,Gibco,+1%BSA,PAA).CellsuspensionsweretreatedwithDHE usingtwicedistilledwater.Therhodiumcontentofthesampleswas (dihydroethidium,Sigma,5μlof5mMstocksolutionper1mlofcell determinedbyatomicabsorptionspectroscopy(AAS,seebelow)and suspensioncontaining106cells)atroomtemperatureinthedarkfor theproteincontentofseparatealiquotsbytheBradfordmethod.Priorto 15min, washed one more time with FACS buffer and immediately AASanalysis20μltritonX-100(1%)and20μlnitricacid(13%)were analyzed using FACS® Calibur (Becton Dickinson) and the Cell addedtoeach200μlsampleofthecellsuspensions.Cellularuptakewas Quest™Proanalysissoftware.Excitationandemissionsettingswere expressedasngrhodiumpermgcellproteinfordataobtainedfrom2 488and564–606nm(FL2filter),respectively.Theregionsof“low” independentexperiments.AASmeasurementswereperformedusinga and“high”ROSinthehistogramplotsweredeterminedaccordingto contrAA 700 high resolution continuum source atomic absorption thecorrespondingcontrolsamples. spectrometer(HRCS-AAS)based on matrixmatched calibrationwith complex 5 as reference standard. Other analytical parameters were 4.8.Mitochondrialoxygenconsumption essentiallyasdescribedrecently[12]. 4.8.1.Isolationofmouselivermitochondria 4.5.Cellularmetabolism Mouse (wildtype, C57BL/6) liver was homogenized by Dounce homogenizationandmitochondriawereisolatedbydifferentialcentri- Changes in cellular metabolism were analyzed using a Bionas fugation in accordance with described procedures [49,50]. The entire 2500 Sensor chip system (Bionas, Rostock, Germany). The sensor isolation procedure was performed in an isolation buffer (300mM chipenablescontinuousmeasurementofoxygenconsumptionusing trehalose,10mMHEPES-KOHpH7.7,10mMKCl,1mMEGTA,1mM Clark-type oxygen-sensitive electrodes [28], pH changes of the EDTA,0.1%fatty acid-free BSA). The homogenatewas centrifuged for mediumbyemployingion-sensitivefieldeffecttransistors[29]and 5minat1000×gand4°C.Supernatantwascollected,transferredinto theimpedancebetweentwointerdigitatedelectrodestructures[27] Eppendorf tubes and centrifuged for 2min at 15800rpm and 4°C. toregistertheimpedanceunderandacrossthecelllayeronthechip Mitochondrialpelletswereresuspendedin1mlofisolationbufferand surface.Beforemeasurement,cellswereseededonthesensorchip thelastcentrifugationstepwasrepeated.Afterresuspendingthefinal (SC 1000) in DMEM (PAA, E15-883) with penicillin/streptomycin mitochondriapelletsin1mlisolationbuffer,theproteincontentwas and10%(v/v)FCS(PAA)andincubatedinastandardtissueculture estimatedbytheBradfordassay(Sigma-Aldrich)at595nm. F.Hackenbergetal./JournalofInorganicBiochemistry105(2011)991–999 999 4.8.2.Measurementofmitochondrialoxygenconsumption [6] A.C.G.Hotze,E.P.L.vanderGeer,H.Kooijman,A.L.Spek,J.G.Haasnot,J.Reedijk, Eur.J.Inorg.Chem.(2005)2648–2657. The measurement was performed using Oxoplate® (PreSens, [7] D.-L. Ma, C.-M. Che, F.-N. Siu, M. Yang, K.-Y. Wong, Inorg. Chem. 46 (2007) Germany), 96-well plates which contain an immobilized oxygen 740–749. sensor at the bottom of each well. Fluorescence was measured in [8] G.I.Pascu,A.C.G.Hotze,C.Sanchez-Cano,B.M.Kariuki,M.J.Hannon,Angew.Chem. Int.Ed.46(2007)4374–4378. dual mode, with excitation at 540nm, emission at 650nm and [9] M.A.Scharwitz,I.Ott,Y.Geldmacher,R.Gust,W.S.Sheldrick,J.Organomet.Chem. reference emission at 590nm. Their signal ratio 650/590nm corre- 693(2008)2299–2309. spondstotheoxygenpartialpressure.Thecalibrationofthefluores- [10] V.I.Sokol,M.A.Porai-Koshits,Sov.J.Coord.Chem.1(1975)577–583. cencereaderwasperformedusingatwo-pointcalibrationwithoxygen- [11] B.R.James,R.H.Morris,Can.J.Chem.58(1980)399–408. [12] M.Harlos,I.Ott,R.Gust,H.Alborzinia,S.Wölfl,A.Kromm,W.S.Sheldrick,J.Med. free water (1% Na 2 SO 3 ) and air-saturated water with oxygen partial Chem.51(2008)3924–3933. pressurecorrespondingto0%and100%respectively.25μgoffreshly [13] X.-N.Li,Z.-J.Wu,Z.-J.Si,L.-Z.Liu,X.-J.Liu,H.-J.Zhang,Phys.Chem.Chem.Phys.11 isolatedmitochondriaweresuspendedin100μl ofrespirationbuffer (2009)9687–9695. [14] D.Herebian,W.S.Sheldrick,J.Chem.Soc.DaltonTrans.(2002)966–974. (25mMsucrose,100mMKCl,75mMmannitol,5mMMgCl ,10mM 2 [15] A. Frodl, D. Herebian, W.S. Sheldrick, J. Chem. Soc. Dalton Trans. (2002) KH 2 PO 4 ,0.5mMEDTA,10mMTris,0.1%fatty acid-freeBSA,pH7.4) 3664–3673. containing10mMpyruvate,2mMmalate,2mMADPand0.5mMATP [16] D.M. Gray, in: G.D. Fasman (Ed.), Circular Dichroism and the Conformational AnalysisofBiomolecules,PlenumPress,NewYork,1996,pp.469–500. toactivateoxidativephosphorylation.Testingsubstanceswerediluted [17] S.Schäfer,W.S.Sheldrick,J.Organomet.Chem.692(2007)1300–1309. sothattheappropriateconcentrationinthewellswasestablishedafter [18] S.Schäfer,I.Ott,R.Gust,W.S.Sheldrick,Eur.J.Inorg.Chem.(2007)3034–3046. addingthemitochondria.Fluorescencewasmeasuredcontinuouslyfor [19] S.Mahadevan,M.Palaniandavar,Inorg.Chem.37(1998)3927–3934. 360minwithintervalsof5minbyaTecanSafire2(Tecan,Männedorf, [ [ 2 2 0 1 ] ] F M .M .E . r P ik o s h s l o , n T. , M B. . N Jo o v r i d n e , n J. , M M o e l t . h B o io d l s .6 En 7 z ( y 1 m 97 o 2 l. ) 3 3 4 7 0 5 ( – 2 3 0 9 0 6 1 . )68–98. Switzerland)microplatereaderat37°C.Duringthemeasurementsthe [22] G.Cohen,H.Eisenberg,Bioploymers4(1996)429–440. platesweresealedwithabreathablemembrane(DiversifiedBiotech, [23] D.Suh,J.B.Chaires,Bioorg.Med.Chem.3(1995)723–728. Boston,MA).Additionalcontrolswere5μMrotenone(Sigma-Aldrich) [24] M.E.Hogan,N.Dattagupta,D.M.Crowthers,Biochemistry18(1979)280–288. [25] Y. Geldmacher, I. Kitanovic, H. Alborzinia, K. Bergerhoff, R. Rubbiani, P. asinhibitor ofrespiratorychain complex I and1μM CCCP (carbonyl Wefelmeier,A.Prokop,I.Ott,S.Wölfl,W.S.Sheldrick,ChemMedChem6(2011) cyanide 3-chlorophenylhydrazone, Sigma-Aldrich) as an uncoupling 1023–1031. agent,capableofincreasingelectronflowthroughtherespiratorychain, [26] A.R.Ghezzi,M.Aceto,C.Cassino,E.Gabano,D.Osella,J.Inorg.Biochem.98(2004) 501–509. therebyincreasingtheoxygenconsumption. [27] R. Ehret, W. Baumann, M. Brischwein, A. Schwinde, B. Wolf, Med. Biol. Eng. Comput.36(1998)365–370. 5.Abbreviations [28] B.Wolf,M.Brischwein,W.Baumann,R.Ehret,M.Kraus,Biosens.Bioelectron.13 (1998)501–509. [29] M.Lehmann,W.Baumann,M.Brischwein,R.Ehret,M.Kraus,A.Schwinde,M. bpy 2,2′-bipyridyl Bitzehofer,I.Freud,B.Wolf,Biosens.Bioelectron.15(2000)117–124. Phen 1,10-phenanthroline [30] R.Rubbiani,I.Kitanovic,H.Alborzinia,S.Can,A.Kitanovic,L.A.Onambele,M. Stefanopoulou,Y.Geldmacher,W.S.Sheldrick,G.Wolber,A.Prokop,S.Wölfl,I. Chrysi 5,6-chrysenquinonediimine Ott,J.Med.Chem.53(2010)8608–8618. dppz dipyrido[2,3-a:2′,3′-c]phenazine [31] M.Dobroschke,Y.Geldmacher,I.Ott,M.Harlos,L.Kater,L.Wagner,R.Gust,W.S. dppn benzo[i]dipyrido[3,2-a:2′,3′-c]phenazine Sheldrick,A.Prokop,ChemMedChem4(2009)177–187. [32] Y.Geldmacher,R.Rubbiani,P.Wefelmeier,A.Prokop,I.Ott,W.S.Sheldrick,J. 2-S-py pyridine-2-thiolate Organomet.Chem.696(2011)1023–1031. 5,6-Me 2 phen 5,6-dimethyl-1,10-phenanthroline [33] R.Bieda,I.Kitanovic,H.Alborzinia,A.Meyer,I.Ott,S.Wölfl,W.S.Sheldrick,in Dpq dipyrido[3,2-d:2′,3′-f]quinoxaline press.Biometals.doi:10.1007/s10534-011-9414-9. [34] V.A.Fadok,D.Xue,P.Henson,CellDeathDiffer.8(2001)582–589. LSIMS liquidsecondaryionmassspectrometry [35] R.A.Schlegel,P.Williamson,CellDeathDiffer.8(2001)551–563. ICD inducedcirculardichroism [36] I. Vermes, C. Haanen, H. Steffens-Nakken, C. Reutelingsperger, J. Immunol. ROS reactiveoxygenspecies Methods184(1995)39–51. [37] C. Batandier, E. Fontaine, C. Keriel, X.M. Leverve, J. Cell. Mol. Med. 6 (2002) AAS atomicabsorptionspectroscopy 175–187. [38] H.-U.Simon,A.Haj-Yehia,F.Levi-Schäfer,Apoptosis5(2000)415–418. [39] P.A.Sandstrom,M.D.Mannie,T.M.Buttke,J.Leukoc.Biol.55(1994)221–226. [40] R.B.P.Elmes,M.Erby,S.M.Cloonan,S.J.Quinn,D.C.Williams,T.Gunnlaugsson, Acknowledgements Chem.Commun.47(2011)686–688. [41] J.G.Collins,A.D.Sleemann,J.R.Aldrich-Wright,I.Greguric,T.W.Hambley,Inorg. Financial support of this work in Bochum, Braunschweig and Chem.37(1998)3133–3141. [42] A.Delgadillo,P.Romo,A.M.Leiva,B.Loeb,Helv.Chim.Acta86(2003)2110–2120. Heidelberg by the Deutsche Forschungsgemeinschaft (DFG) within [43] M.J.Frisch,G.W.Trucks,H.B.Schlegel,G.E.Scuseria,etal.,Gaussian03,Revision the research group FOR 630 “Biological Function of Organometallic E.01,GuassianInc.,WallingfordCT,2004. Compounds”isgratefullyacknowledged. [44] D.Feller,J.Comput.Chem.17(1996)1571–1586. [45] K.L.Schuchardt,B.T.Didier,T.Elsethagen,L.S.Sun,V.Gurumoorthi,J.Chase,J.Li,T.L. Windus,J.Chem.Inf.Model.47(2007)1045–1052. References [46] A.-R.Allouche,J.Comput.Chem.32(2011)174–182. [47] J.Marmur,J.Mol.Biol.3(1961)208–218. [1] K.E.Erkkila,D.T.Odom,J.K.Barton,Chem.Rev.99(1999)2777–2795. [48] I.Ott,K.Schmidt,B.Kircher,P.Schuhmacher,T.Wiglenda,R.Gust,J.Med.Chem. [2] C.Metcalfe,J.A.Thomas,Chem.Soc.Rev.32(2003)215–224. 48(2005)622–629. [3] N.J.Wheate,C.R.Brodie,J.G.Collins,S.Kemp,J.R.Aldrich-Wright,MiniRev.Med. [49] E. Fernandez-Vizarra, M.J. Lopez-Perez, J.A. Enriquez, Methods 26 (2002) Chem.7(2007)627–648. 292–297. [4] B.A.Jackson,J.K.Barton,J.Am.Chem.Soc.119(1997)12986–12987. [50] R. Yamaguchi, A. Andreyez, A.N. Murphy, G.A. Perkins, M.H. Ellisman, D.D. [5] U.Schatzschneider,J.Niesel,I.Ott,R.Gust,H.Alborzinia,S.Wölfl,ChemMedChem Newmeyer,CellDeathDiffer.14(2007)616–624. 3(2008)1104–1109.