A potential method to improve the in vitro cytotoxicity of half-sandwich Os(ii) complexes against A2780 cells.

Dalton Transactions PAPER in vitro A potential method to improve the cytotoxicity of half-sandwich Os( ) complexes II Citethis:DOI:10.1039/c8dt00193f † against A2780 cells PavelŠtarha, aZdeněkTrávníček, *aRadovanHerchel, aPawelJewulaaand ZdeněkDvořák b The[Os(η6-pcym)(dpa)(VP)]PF (1-VP)complexcontainsthehistonedeacetylase(HDAC)inhibitorvalpro- 6 ate(2-propylpentanoate;VP)asamonodentateO-donorligandandshowsca.3-foldhigherinvitrocyto- toxicityagainstA2780humanovariancarcinomacellsthanitschloridoanalogue[Os(η6-pcym)(dpa)Cl]PF 6 (1-Cl); pcym= 1-methyl-4-(propan-2-yl)benzene(p-cymene), dpa=2,2’-dipyridylamine. Thecomplex 1-VPshowedpromisingselectivitytowardstheA2780ovariancarcinomacellline(IC =20.9µM)over 50 normal human hepatocytes (IC > 200.0 µM). Moreover, the complex 1-VP was found to be inactive 50 againstMCF-7(breastadenocarcinoma),PANC-1(pancreaticadenocarcinoma)andHT-29(coloncarci- noma) up to a concentration of 100 µM. Detailed flow cytometry studies indicated that treatment of A2780 cells with complex 1-VP led to induction of apoptosis, production of reactive oxygen species (ROS)andsuperoxide(SO)anionradicals,aswellasmitochondrialmembranepotentialdepletionandcell cycleperturbations.Themicroscopicassessment(standardhematoxylin/eosinstaining)revealedsignsof morphologicalchangesassociatedwiththeprogressionofapoptosisinA2780cellstreatedwiththeIC 50 concentration of the complex 1-VP. Consistent with the intracellular production of ROS and SO, the Received16thJanuary2018, complex1-VPinducedhydroxylradicalformation,asprovedbyEPRspintrappingexperiments.Thiscase Accepted20thMarch2018 studysuggeststhatreplacementofthechloridoligandofhalf-sandwichOs(II)complexesbyareleasable DOI:10.1039/c8dt00193f monodentatebiologicallyactiveligand(e.g.,VPusedinthisstudy)isaneffectivestrategyforthedevelop- rsc.li/dalton mentofnovelnon-platinumcytotoxicagents. Introduction that of conventional platinum drugs and comprises multiple effectssincerelevantevidencehasbeenreportedfortheinvol- Toovercomethedrawbacksofplatinum-basedchemotherapy,1 vement of their interaction with DNA4 as well as redox- complexesofother transitionmetalshave beenstudied inten- mediatedprocesses.3a,5 sively for anticancer activity.2 Among these, half-sandwich Theanticancerpotencyofthe[Os(η6-ar)(L )X]0/+complexes n [Os(η6-ar)(L )X]0/+ complexes(ar=arene, e.g.,p-cymene,pcym, can be tuned by replacement of the halogenido ligand by a n L =abidentateligandandX=ahalogenidoligand),2b,3such bioactive ligand such as organic inhibitors of various n as highly cytotoxic [Os(η6-pcym)(L )I]PF ; L = N,N-dimethyl-4- enzymes.Asimilarstrategywasrecentlyusedforthedevelop- 1 6 1 [(E)-pyridin-2-yl-diazenyl]-aniline,3a appear to be a pharma- ment of dual-acting (bifunctional) platinum(IV) complexes cologicallyprospectivegroupofcompounds.Theirmechanism containing various bioactive axial O-donor ligands (e.g., ofaction,althoughstillnotfullyunderstood,isdifferentfrom valproate or 4-phenylbutyrate).6 Regarding Os(II) complexes, this approach has been used only for the complex [Os(η6- pcym)(acac)(EC)] containing the ethacrynate anion (EC): acac=pentane-2,4-dionate.7Ethacrynicacid(ECA)isamono- aDepartmentofInorganicChemistry&RegionalCentreofAdvancedTechnologies carboxylic acid that acts as an inhibitor of glutathione andMaterials,FacultyofScience,PalackýUniversityinOlomouc,17.listopadu12, 77146Olomouc,CzechRepublic.E-mail:zdenek.travnicek@upol.cz S-transferases (GSTs)8 responsible for the conjugation of bDepartmentofCellBiologyandGenetics&RegionalCentreofAdvanced reducedglutathione(GSH)incytosolwithvariousxenobiotics TechnologiesandMaterials,FacultyofScience,PalackýUniversityinOlomouc, including metal-based anticancer agents. However, the cyto- Šlechtitelů27,78371Olomouc,CzechRepublic toxicity of [Os(η6-pcym)(acac)(EC)] (IC = 43 µM) did not †Electronic supplementary information (ESI) available: Experimental section, 50 TablesS1–S6andFig.S1–S13.CCDC1515628.ForESIandcrystallographicdata exceed that of free ECA (IC 50 = 40 µM); this implied that the inCIForotherelectronicformatseeDOI:10.1039/c8dt00193f complex acted only as a drug carrier for the bioactive EC Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online View Journal Paper DaltonTransactions ligand,whichpredominantlyprovokedtheresultingcytotoxic The 1H and 13C NMR spectra of the complex 1-VP con- effects. tained, in addition to the pcym and dpa signals also detected Inthisstudy,weusedvalproatemonoanion(VP)asamono- inthespectraofthecomplex1-Cl,thesignalsoftheVPligand dentate O-donor bioactive ligand of the innovative Os(II) (Fig.1,S1andS2,ESI†).Forexample,thecharacteristictriplet complex [Os(η6-pcym)(dpa)(VP)]PF (1-VP; dpa = 2,2′-dipyridyl- of the terminal methyl groups of VP (i.e., C25–H ) was 6 3 amine; Fig. 1). Valproic acid (VPA) is approved for the treat- observed at δ = 0.63 ppm in the 1H NMR spectrum of the ment of seizures and bipolar disorder.9 VPA has also been complex1-VP.Thisδvaluediffersmarkedlyfromδ=0.85ppm investigatedasapossibleantitumorsubstance,actingthrough of free VP (measured on NaVP) and correlated well with those histone deacetylase (HDAC) inhibition.10 The hyperacetylated reported for the complexes [Ru(η6-pcym)(dpa)(VP)]PF (δ = 6 histonescannotbeassociatedwithDNA,whichbecomesmore 0.62 ppm)13 and trans-[Pt(py) (VP) ] (δ = 0.65 ppm)14 contain- 2 2 accessibletoDNA-attackingagents11suchasmetal-basedanti- ing VPas a monodentate O-donor ligand. The replacement of cancer agents, including Os(II) complexes.4 To the best of our the chlorido ligand by the VP ligand resulted in δ changes in knowledge, the effects of a releasable HDAC inhibitor on the the1HNMRsignalsofpcymanddpa(Fig.1).Forexample,the cytotoxicity of half-sandwich Os(II) complexes have not yet characteristic C6–H signal of dpa as well as the signals of the beenreportedintheliterature. pcym aromatic hydrogens (C12, C13, C15, C16) shifted down- field by ca. 0.1 ppm. The 13C NMR spectra contained the characteristicsignaloftheC21quaternarycarbonatomofthe carboxylgroupoftheVPligand(Fig.S1,ESI†),whichisunam- Results and discussion biguously assignable from the 1H–13C gs-HMBC spectra con- Synthesisandcharacterization taining the C21 (δ = 181.2 ppm) ↔ H23 (δ = 1.23 ppm) long- The complex [Os(η6-pcym)(dpa)Cl]PF (1-Cl; Fig. 1) was pre- rangecorrelationpeak.Nonewsignalsweredetectedinthe1H 6 and13CNMRspectraofthecomplexes1-VPand1-Cldissolved paredbyamodificationoftheprocedurereportedrecentlyfor its analogue [Os(η6-bz)(dpa)Cl]PF (bz = benzene).12 The in DMSO-d 6 after 48 h of standing at ambient temperature, 6 suggestingtheirstabilityintheusedsolvent. complex1-VP(Fig.1)wassynthesizedinreasonableyieldfrom The ESI+ mass spectra of the complexes 1-VP and 1-Cl con- the complex 1-Cl by reacting it with an excess of silver valpro- ate, as recently reported for the Ru(II) analogue [Ru(η6-pcym) tained the peaks of the [Os(pcym)(dpa)(VP)]+ (639.2 m/z; calc. (dpa)(VP)]PF .13Thecomplex1-VPrepresentsthefirstosmium 639.6,10%relativeintensity)and[Os(pcym)(dpa)Cl]+(532.1m/z; 6 calc. 532.1, 65% relative intensity) species, respectively (Fig. 2). valproatocomplexaswellasthefirsthalf-sandwichosmium(II) TheusedionizationconditionsledtoreleaseoftheVP(for1-VP) complex containing an HDAC inhibitor as a releasable mono- andchlorido (for1-Cl) ligands and formation of the {[Os(pcym) dentateO-donorligand. (dpa)]–H}+species(496.2m/z;calc.496.2,100%relativeintensity) (Fig.2).TheresultsoftheRP-HPLC/ESI+MSexperimentsproved the>99.5%purityofthecomplexes1-VPand1-Clandindicated the higher relative hydrophobicity15 of the complex 1-VP (t = R 20.2 min) compared with the complex 1-Cl (t = 13.7 min). The R presence of the VP ligand in the structure of the complex 1-VP wasalsoprovedbyFTIRspectroscopy(Fig.S3,ESI†). SinglecrystalX-rayanalysis Attempts to prepare single crystals of the complexes 1-Cl and 1-VPsuitableforsingle-crystalX-rayanalysiswereunsuccessful; Fig.1 Structural formulas of the complexes [Os(η6-pcym)(dpa)Cl]PF Fig.2 ESI+ mass spectra of the complexes 1-VP (top) and 1-Cl 6 (1-Cl)and[Os(η6-pcym)(dpa)(VP)]PF (1-VP)withtheiratomnumbering (bottom) given with the experimental (inset in black) and simulated 6 schemes(top),andtheir1HNMRspectrawithgeneralassignmentofthe (inset in grey) isotopic distributions of the [Os(pcym)(dpa)(VP)]+ (for observedsignals(bottom). 1-VP)and[Os(pcym)(dpa)Cl]+(for1-Cl)species. DaltonTrans. Thisjournalis©TheRoyalSocietyofChemistry2018 .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online DaltonTransactions Paper however, the exchange of PF − with BPh − (see Experimental [Ru(η6-pcym)(dpa)Cl]PF ;inparticular,C14inthetext=C11in 6 4 6 section) led to the formation of crystals of the complex theCIFfile;C20inthetext=C27intheCIFfile.17 [Os(η6-pcym)(dpa)Cl]BPh ·CH OH(1-Cl′·CH OH),whichallowed The dihedral angle between both pyridine rings of dpa in 4 3 3 us to determine the structure of the [Os(η6-pcym)(dpa)Cl]+ thecomplex1′·CH OHequals39.24(11)°.Thepcymringforms 3 complex cation. The complex 1-Cl′·CH OH adopts a half- dihedralanglesof28.28(8)°(withthepyridineringcontaining 3 sandwich geometry with the bidentate N,N-donor dpa ligand, theN1atom)and34.32(11)°(withthepyridineringcontaining η6-coordinated pcym and chlorido ligand in the inner coordi- the N1A atom). N–H⋯O, C–H⋯C, O–H⋯π and C⋯C non- nationsphere(Fig.3,S4andS5,TablesS1andS2,ESI†). covalent contacts were detected in the crystal structure of the Todate,onlytenX-raystructuresofmononuclearhalf-sand- complex 1′·CH OH (Table S2 and Fig. S5, ESI†). The shortest 3 wich p-cymene-chloridoosmium(II) complexes containing a Os⋯B (towards BPh 4 − counterion)and Os⋯O (towards MeOH bidentate N-donor ligand coordinated through two nitrogen molecule of crystallization) distances are 6.508(3) Å and heteroatoms have been deposited in the Cambridge 6.026(2)Å,respectively. Crystallographic Data Centre (CCDC) under the respective Cambridge Structural Database (CSD) refcodes, CSD version StudiesofhydrolysisandinteractionswithGSH 5.37, updated May 2017.16 For these complexes, which are The complex 1-Cl hydrolysed in a 10% MeOD-d /90% D O 4 2 structurally similar to the herein reported complex 1- mixture, as evidenced by the detection of new signals in the Cl′·CH 3 OH, the average values of the Os–Cl and Os–N bond 1H NMR spectrum which are assignable to the [Os(η6-pcym) lengthsare2.401(12)Å(2.383to2.431Å),and2.09(3)Å(2.035 (dpa)(OD )]2+ hydrolysate(1-OD ;Fig.S6,ESI†).Approximately 2 2 to2.148Å),respectively,whichcorrelateswiththevaluesdeter- 50% of the complex 1-Cl hydrolysed after 48 h of standing at mined for the complex 1-Cl′·CH 3 OH (Fig. 3). Among the ambient temperature. The hydrolysis progress reversed after reported structures, the bond lengths of Os–Cl = 2.402(2) Å the addition of KCl, resulting in the disappearance of the and Os–N = 2.096(6) and 2.101(6) Å of the complex [Os(η6-bz) signalsofthe1-OD speciesandtheformationofthecomplex 2 (dpa)Cl]PF 6 (CCDC 1505445;† the benzene analogue of the 1-Cl. pcym complex 1-Cl)12 were comparable with those of the Similarly, the complex 1-VP hydrolysed in the same complex 1-Cl′·CH 3 OH. Concerning the complex [Ru(η6-pcym) medium, and signals assignable to the 1-OD 2 species and (dpa)Cl]PF 6 (CSDref code:IKEKUF; CCDC701724†),17aRu(II) released (free) VP were detected (Fig. 4). In particular, the 1H analogue of the complex 1-Cl, both complexes have different NMRspectrumrecordedforafreshsolution ofthemixture of orientationsofthepcymringtowardstheM–Clbond(Fig.S4, thecomplex1-VPin10%MeOD-d /90%D Ocontainedtwotri- 4 2 ESI†). In particular, the values of the Cl–M–C14 bond angle andCl–M–C14–C20torsionangleequal95.78(8)°and33.5(3)°, respectively, for the complex 1′·CH OH; these differ markedly 3 from the values of 155.00(6)° and 101.2(2)°, respectively, reported for its Ru(II) analogue. Note: Different labels from the CIF file (CCDC 701724†) were used for the complex Fig.3 Molecularstructureofthecomplexcationof1-Cl’·CH OHwith 3 thenon-hydrogenatomsdrawnasthermalellipsoidsatthe50%prob- abilitylevel.TheBPh −counterionandCH OHmoleculeofcrystalliza- 4 3 tion were omitted forclarity (CCDC 1515628†). Selected bond lengths Fig.4 Top:1HNMRstudiesofthecomplex1-VPin10%MeOD-d /90% 4 (Å) and angles (°): Os–Cl = 2.3923(8), Os–N1 = 2.109(3), Os–N1A = D O,depictingthesignalsoftheC25–H groupsofcoordinated(blue) 2 3 2.095(4), Os–Cg = 1.6764(4), Os–C11 = 2.210(3), Os–C12 = 2.200(4), orfree(red)valproate.The1HNMRspectrumofNaVPisgivenforcom- Os–C13 = 2.197(3), Os–C14 = 2.216(3), Os–C15 = 2.193(3), Os–C16 = parative purposes. Bottom: Summary of the 1H NMR studies of the 2.201(3), Cl–Os–N1 = 84.86(9), Cl–Os–N1A = 84.91(9), Cl–Os–Cg = complex1-VPin10%MeOD-d /90%D O,givenastheextentoftheVP 4 2 127.92(2), N1–Os–N1A = 82.41(12), N1–Os–Cg = 130.21(8), N1A–Os– ligandrelease(%)underdifferentexperimentalconditionsanddetected Cg=130.00(10).Cg=thecentroidofthepcymaromaticringdefined atdifferenttimepoints.+and−symbolizethepresenceandabsence, bytheC11–C16atoms. respectively,ofPBS,GSHand0.1%KODinthestudiedmixtures. Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online Paper DaltonTransactions plets of the C25–H hydrogens of coordinated (δ = 0.62 ppm) 3 andfreevalproate(δ=0.78ppm;correlatedwithδ=0.78ppm detectedforNaVPdissolvedinthesamemedium)(Fig.4).The ratios detected for the released and coordinated VP were ca. 15:85(freshsolution,t=0h)andca.35:75after6hofstand- ing at ambient temperature, when the equilibrium state was reached. In phosphate buffer (pH 7.2), the complex 1-VP released its VP ligand at a higher rate (55% in the fresh solu- tion)andtoagreaterextent (85%after6h)(Fig.4).Theeffect of the chloride ions present in the used buffer must be taken intoaccount,assuggestedbythepartialtransformationofthe Fig.5 A comparison of the obtained IC values (µM; given together 50 formed aqua species 1-OD to the chlorido species (Fig. S7, 2 with SD) against A2780 cells of the complexes 1-VP and 1-Cl (** for ESI†). p<0.01). Aslightlylowerextent ofVP ligandreleasewas observedin the presence of 2 mol equiv. of the intracellular tripeptide GSH(Fig.4).Theseexperimentswereperformedinphosphate KP1019,whichhasRA=0.12.24Note:theRAvaluesweredeter- buffer (pH 7.2) and in 10% MeOD-d 4 /90% D 2 O with the mined for A27807,21,22 or CH13c,23 human ovarian carcinoma addition of 0.1% KOD (pH 7.2). The extent of the VP ligand cellsagainstthereferencedrugcisplatin;L =N,N-dimethyl-N′- 2 release reached 75% and 80% for these two experiments, [(E)-pyridin-2-ylmethylidene]benzene-1,4-diamine. respectively(Fig.4).Interestingly,althoughno1HNMRsignals In contrast to A2780 cells, the MCF-7 breast, PANC-1 pan- assignable to the covalent adducts of the complex 1-VP with creatic and HT-29 colon carcinoma cell lines were markedly GSH were detected in the phosphate buffer solution, new 1H less sensitive to the cytotoxic action of complex 1-VP, and the NMR signals assignable to the [Os(η6-pcym)(dpa)(SG)]+ adduct IC valueswerenotreachedinthetestedconcentrationrange 50 (1-SG) clearly showed in the spectra obtained for the mixture (IC >100.0µM). 50 of1-VPandGSHin10%MeOD-d 4 /90%D 2 O,whosepHwasset Thetoxiceffectsofthecomplex1-VPwerenegligible(IC 50 > to 7.2 using 0.1% KOD in D 2 O (Fig. S8, ESI†). Specifically, the 200.0 µM) for the studied primary culture of healthy human signals of the 1-SG adduct appeared at 2.13 ppm (Cys-β-CH 2 ) hepatocytes, which is indicative of pharmacologically promis- and 3.50 ppm (Cys-α-CH). Formation of the 1-SG species was ing selectivity of the complex 1-VP towards A2780 cells over proved by ESI− mass spectrometry (found: 801.1 m/z, calc.: normalcells. 801.2m/z;Fig.S9,ESI†). Cellularaccumulation Invitrocytotoxicity Althoughcellularuptakebycancercellsisacrucialmechanis- It is known that the cytotoxicity of half-sandwich osmium(II) tic aspect in the anticancer action of xenobiotics,25 its rate complexes can be enhanced by replacement of the chlorido does not directly determine cytotoxicity. In many cases, less ligand with an iodido3a or bromido18 ligand. In this work, we accumulated substances exceed the potency of analogues dis- replaced the chlorido ligand with a known HDAC inhibitor playinghighercellularuptakes,asalsoreportedforhalf-sand- (i.e.,VP).10,11 The complex 1-Cl was moderately effective against A2780 wich Os(II) complexes [Os(η6-ar)(npic)Cl]; ar = e.g., pcym or biphenyl, npic = 2-picolinate or 4-methyl-2-picolinate.4,20 cells (59.9 ± 5.8 µM). A significant increase (p < 0.01) of cyto- Similarly, the difference in cytotoxicity of the complexes 1-VP toxicity against A2780 cells resulted from replacement of the and 1-Cl (Fig. 5) cannot be associated with cellular accumu- chloridoligandofthecomplex1-ClbyVPtoformthestructure lation because the less potent complex 1-Cl is accumulated of the complex 1-VP, which displayed ca. 3-fold higher cyto- more (280.1 ± 8.8 ng Os per 106 cells) than the complex 1-VP toxicity(IC =20.9±1.9µM)thanitsprecursor1-Cl(Fig.5). 50 (236.6±19.3ngOsper106cells). Bothcomplexes1-Cland1-VPexceededtheinvitrocytotoxi- cities of their Ru(II) analogues (IC 50 > 100 µM).13 The cyto- toxicity of free VP against A2780 cells (IC = 7000 µM)19 was Cellcycleanalysis 50 more than two orders of magnitude lower than that of the The complexes 1-VP and 1-Cl (and cisplatin for comparative complex 1-VP. The relative activity (RA; defined as purposes) were applied to A2780 cells at their IC concen- 50 IC (cisplatin)/IC (complex) toward A2780 cells) equalled ca. trations, and the DNAcontentwas assessedby flow cytometry 50 50 0.34 for the complex 1-VP. Although the RA of the complex 1- usingconventionalpropidiumiodide(PI)staining.Theresults VP is lower than that of one of the best-performinghalf-sand- aregiveninFig.6AandinTableS3(ESI†). wich Os(II) complexes, [Os(η6-pcym)(L 1 )I]PF 6 (RA = 10),3a it is The complexes 1-VP and 1-Cl induced negligible modifi- comparable with those of other Os(II) analogues,5c,20 such as cation of the A2780 cell cycle compared with the negative [Os(η6-pcym)(L )Cl]PF (RA=0.40),andevenhigherthanthose controlbecausetheobservedpopulationsforthecomplexes1- 2 6 of many half-sandwichOs(II) agents(e.g., ref. 3c, 7 and 21–23) VP and 1-Cl are comparable with that of the negative control. as well as clinically investigated half-sandwich Ru(II) complex Onlyastatisticallysignificantbutbiologicallyirrelevantdiffer- DaltonTrans. Thisjournalis©TheRoyalSocietyofChemistry2018 .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online DaltonTransactions Paper induced by the complexes 1-Cl and 1-VP appears to be different from that of the complex [Os(η6-pcym)(L )I]PF , 1 6 which induces A2780cancercelldeaththroughanon-apopto- ticmechanism.5a Cellmorphologystudies Theabilityofcomplex1-VP(appliedat20.9μMconcentration) to induce cell death in A2780 cells was also confirmed by microscopic observation (after standard hematoxylin/eosin staining) of morphological changes associated with the pro- gression of apoptosis (e.g., chromatin condensation, microsa- tellite formation, cell shape changes, vacuolization and destruction of cell membranes). However, the effects of the complex1-VPonthemorphologyofA2780cellsweremoderate compared with those of cisplatin (applied at 10 μM concen- tration), which acts as a strong cytotoxic agent and apoptosis inducer(Fig.S10,ESI†). Fig.6 (A)Populations(%)ofA2780cellstreatedwiththecomplexes1- ROS/superoxideinduction VPand1-Cl(andcisplatinforcomparativepurposes)withthenegative control (PI/RNase staining). The data are given asthe arithmetic mean Apoptosis, which is mainly mitochondria-mediated, is usually from three independent experiments. (B) Selected results of apoptosis connected with an elevated level of intracellular reactive inductionshownforA2780cellstreatedwiththecomplex1-VPfor2h oxygen species (ROS), which affect the redox equilibrium in (green)and24h(blue)andforthenegativecontrol(red). the treated cancer cells.27 For this reason, the ability of the complexes 1-VP and 1-Cl (IC concentrations) to induce ROS 50 and/orsuperoxide(SO)inA2780cellswasevaluatedbyastan- ence in the G /G -cell cycle phase population of cells treated 0 1 dardassay(Fig.7andTableS5,ESI†). with the complex 1-VP in comparison with cells treated with Applicationof1-Cland1-VPtoA2780cellsledtothesimul- the complex 1-Cl or with the negative control can be men- taneousinductionofbothROSandSObecausethemajorityof tioned here (Fig. 6A). Moreover, the complexes 1-VP and 1-Cl perturbedtheA2780cellcycledifferentlythancisplatin,which the treated A2780cellswere detectedinthe upperright quad- indicates that these agents have a different mechanism of rantofthedotplot (Fig.7).Thisbehaviourissimilartothatof action. The cellcycle modifications inducedby the complexes the recently reported Os(II) complex [Os(η6-pcym)(L 1 )I]PF 6 , whichwasalsostudiedinA2780cells.5a,28 1-VP and 1-Cl are consistent with those of similar half-sand- wichOs(II)complexesstudiedinA2780cells;theyalsoshowed Mitochondrialmembranepotentialassay higherG /G -cellcyclephasepopulationsandlowerS-andG / 0 1 2 M-cellcyclephasepopulationscomparedwithcisplatin.3a,5a ApoptosisandelevatedROS(includingSO)alterthemitochon- drial membrane potential (Δψ ) of treated cells,27,29 as m Inductionofapoptosis reported for various anticancer agents, including cisplatin.30 In this experiment, the combined effects of Annexin V-FITC Because the complexes 1-VP and 1-Cl induced apoptosis and and propidium iodide (PI) markers on cells are used to dis- high levels of ROS (including SO), their effects on Δψ m of tinguishbetweenviablecells,earlyapoptoticcells,lateapopto- A2780cellswerestudiedusingacommercialkit(Fig.8). ticcellsandnecroticcells.Inductionofapoptosiswasstudied TheexposureofA2780cellstothecomplexes1-VPand1-Cl forthecomplexes1-VPand1-Clappliedforeither2hor24h inducedΔψ m changes,resultingindifferentratiosofthecyto- at their IC concentrations to A2780 cells (cisplatin and 50 known apoptosis-inducer staurosporine26 were used for com- parative purposes). The results are given in Fig. 6B and in TableS4(ESI†). The complexes 1-VP and 1-Cl induced apoptosis of A2780 cells. The apoptotic cell populations (78.4% of cells for 1-VP and78.5%ofcellsfor1-Cl)inducedbythestudiedOs(II)com- plexesaremutuallycomparableafter24hoftreatmentandare markedlyhigherthanthatofA2780cellstreatedwithanequi- potentconcentrationofcisplatin(56.1%ofcells).Theonsetof Fig.7 Cell populations observed by flow cytometry studies after the apoptosis (studied after 2 h exposure time) was markedly inductionofROSandsuperoxidebythecomplexes1-Cl(blue)and1-VP slowerforthevalproatocomplex1-VP(50.8%ofcells)thanfor (green); negative (red) and positive (pyocyanin; orange) controls are the precursor 1-Cl (64.1% of cells). The apoptotic cell death givenforcomparativepurposes. Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online Paper DaltonTransactions Fig.9 Results of the EPR spin trapping experiments, showing the Fig.8 (A)Mitochondriamembranepotentialflowcytometryanalysisof signalsoftheDEPMPO/OHadductsdetectedinthesystemscontaining the complexes 1-VP or 1-Cl mixed with H O and the DEPMPO spin untreated A2780 cells (control; red) and A2780 cells treated with the 2 2 trap. complexes1-VP(blue),1-Cl(pink)andCCCP(positivecontrol;green), and (B) summary of the A2780 cell populations (%) detected by flow cytometry in the high-orange fluorescence area for untreated A2780 cells (control) and A2780 cells treated with the complexes 1-VP, 1-Cl, The experimental EPR spectra were fitted using EasySpin cisplatinandCCCP.**forp<0.01and***forp<0.005. software33 with two defined spin subsystems for S = 1/2 and varyingg ,A (N),A (P)andA (H).Thevaluesofthehyper- iso iso iso iso finesplittingparameterswereusedtoassignaparticularspin solic monomer form (green fluorescence; FL1) and intramito- system to either cis- or trans-DEPMPO/OH radical diastereo- chondrialorange-fluorescentaggregates(FL2)oftheuseddye. isomer (Fig. S11 and S12, ESI†).34 Based on literature reports, In particular, significant mitochondrial depolarization the two isomers should have similar values of A (N) and iso induced by the complexes 1-VP and 1-Cl led to shifts of the A (P) but significantly different values of A (H); also, cis- iso iso orange fluorescence to lower intensitiesthan observed for the A (H)>trans-A (H).34Theassignmentwasfurthersupported untreatedcells(Fig. 8,TableS6, ESI†).The populations ofthe iso iso with Density Functional Theory (DFT) calculations performed Δψ m depleted cells were comparable for the complexes 1-VP using ORCA 4.0.0,35 in which the molecular geometries of and 1-Cl (no statisticallysignificant difference); however, both both geometrical isomers were optimized and g-tensors and these complexes induced lower increases of A2780 cell popu- hyperfine A-tensors were computed (Fig. S13, ESI†). First, the lations with depolarized mitochondria than cisplatin applied B3LYP functional36 and def2-TZVP basis set37 for all atoms attheequipotentconcentration. were employed together with the Conductor-like Polarizable Continuum Model (CPCM) for water38 and atom-pairwise dis- Cell-freehistonedeacetylaseassay persion correction to the DFT energy with Becke–Johnson In the used cell-free assay, the complex 1-VP reduced HDAC damping (D3BJ)39 to optimize the geometries of both dia- activityby30.5±2.5%,whichwascomparablewiththeHDAC stereoisomers using TightOpt criteria. The resulting geome- activityinhibitioninducedbyfreeVPA(30.6±2.9%).Forcom- tries are visualized in Fig. S13.† Subsequently, the g-tensors parative purposes, the mixture containing cisplatin was evalu- and hyperfine A-tensors were calculated with the B3LYP func- ated as well, and the obtained results indicated low HDAC tional using the EPR-III basis set for C, N, O, and H atoms40 activity inhibitory efficiency (9.4 ± 2.2%). The half-sandwich anddef2-TZVPPforPatom.Theresultingparametersaresum- Ru(II) complex containing a 1,10-phenanthroline derivative, marizedinFig.S11†forbothisomers.Thecomputedvaluesof mimicking the known HDAC inhibitor suberoylanilide hydro- A (X)(X=N,P,H)areevidentlyshiftedinrespecttothefitted iso xamic acid (SAHA), showed reduced HDAC activity of 98.4%, values from the experimental data, but the trend isthe same: while free SAHA (1 μM concentration) revealed a value of cis-A (N) > trans-A (N), cis-A (P) < trans-A (P) and cis- iso iso iso iso 99.5%.31 A (H)≫trans-A (H). iso iso EPRspin-trappingstudies Inadditiontotheflowcytometricresults,whichdemonstrated Experimental high levels of ROS and SO in the treated A2780 cancer cells, Materials EPR spin-trapping experiments were performed with the 5-diethoxyphosphoryl-5-methyl-1-pyrroline-N-oxide (DEPMPO) The chemicals OsCl ·xH O, 1-methyl-4-(propan-2-yl)cyclohexa- 3 2 spin trap.32 The experimental results proved the formation of 1,3-diene(α-terpinene),2,2′-dipyridylamine(dpa),valproicacid (cid:129) hydroxyl radicals (OH) from hydrogen peroxide in the pres- (VPA), sodium hydroxide, silver trifluoromethanesulfonate, enceofthecomplexes1-VPand1-ClbecauseintensiveX-band potassiumchloride,ammoniumhexafluorophosphate,sodium EPRsignalsoftheDEPMPO/OHadductswereobtainedforthe tetraphenylborate, (SP-4-2)-diamminedichloroplatinum(II) (cis- samples containing the complexes 1-VP and 1-Cl (Fig. 9 and platin), staurosporine, GSH, hydrogen peroxide (30.0 wt% in S11, ESI†). The signals of the DEPMPO/OH adduct were H O), DClO and KOD, solvents (methanol, diethyl ether), 2 4 quenched by the addition of 2 mol equiv. of GSH (Fig. S12, HPLC grade solvents (acetonitrile, water) and deuterated sol- ESI†). vents (DMSO-d , MeOD-d , D O) were supplied by VWR 6 4 2 DaltonTrans. Thisjournalis©TheRoyalSocietyofChemistry2018 .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online DaltonTransactions Paper International (Stříbrná Skalice, Czech Republic), Sigma- C4–H,2H),7.16(d,J=8.4Hz,C3–H,2H),7.07(br,C5–H,2H), Aldrich(Prague,CzechRepublic),FisherScientific(Pardubice, 6.07 (d, J = 5.7 Hz, C13–H, C15–H, 2H), 6.02 (d, J = 5.7 Hz, Czech Republic) and Litolab (Chudobín, Czech Republic). All C12–H,C16–H,2H),2.37(sep,J=6.6Hz,C17–H,1H),2.04(m, the chemicals and solvents were used without further purifi- C22–H,1H),2.01(s,C20–H,3H),1.23(m,C23–H,4H),1.07(d, cation.RoswellParkMemorialInstitute(RPMI-1640)medium, J = 7.1 Hz, C18–H, C19–H, 6H), 0.99 (m, C24–H, 4H), 0.63 (t, trypsin and phosphate-buffered saline (PBS) were purchased J = 7.1 Hz, C25–H, 6H). 13C NMR (DMSO-d , ppm): δ 181.2 6 fromSigma-AldrichandFisherScientific,while5-diethoxypho- (C21),154.1(C6),151.4(C2),140.5(C4),118.7(C5),114.1(C3), sphoryl-5-methyl-1-pyrroline-N-oxide(DEPMPO)waspurchased 93.7(C11),91.1(C14),74.9 (C13,15),74.2(C12,16),46.2(C22), fromEnzoLifeSciences(Farmingdale,NY,USA). 35.3 (C23), 30.6 (C17), 22.2 (C18,19), 20.1 (C24), 17.7 (C20), [Os(µ-Cl)(η6-pcym)Cl] was prepared as reported in the lit- 13.9(C25). 2 erature.41 Silver valproate (AgVP) was prepared utilizing VPA, [Os(η6-pcym)(dpa)(OD )]2+(1-OD ).Astoichiometricamount 2 2 whichwasdissolvedin5mLofMeOHandthenneutralizedby of silvertrifluoromethanesulfonatewas added to asolution of 1 M NaOH (in methanol), followed by the addition of a stoi- 1-Cl in MeOD-d . The mixture was stirred for 1 h under alu- 4 chiometric amount of silver trifluoromethanesulfonate minium foil at ambient temperature. After that, AgCl was (triflate). removed and the obtained solution of the dechlorinated (aquated)species1-OD wasmaintainedinarefrigeratoruntil 2 Syntheses required. [Os(η6-pcym)(dpa)Cl]PF (1-Cl). [Os(µ-Cl)(η6-pcym)Cl] 6 2 (0.5 mmol) was reacted in a microwave reaction system (100 °C, 1 min) with 1.1 mmol of dpa in 5 mL of MeOH. Methods Thereafter, the solution was cooled to ambient temperature; The syntheses of [Os(µ-Cl)(η6-pcym)Cl] , 1-Cl and 1-Cl′ were then, an excess of NH PF (2.5 mmol) was added, and this 2 4 6 carriedoutinaMonowave300microwavereactor(AntonPaar) mixture was stirred without heating for 5 min. The solution using30mLmicrowavevials. was filtered and concentrated until the yellow complex 1-Cl Elementalanalyses(C,H,N)werecarriedoutusingaFlash formed.Theproductwascollectedbyfiltration,washed(1mL 2000 CHNS Elemental Analyser (Thermo Scientific). 1H, 13C, of MeOH and 3 × 2 mL of diethyl ether) and dried under 1H–1H gs-COSY, 1H–13C gs-HMQC and 1H–13C gs-HMBC vacuum. The yield was ca. 65% relative to the starting Os(II) spectra were recorded for DMSO-d solutions using a JEOL dimer. Anal. calcd for OsC H N ClPF : C, 35.53; H, 3.43; N, 6 20 23 3 6 JNM-ECA 600II spectrometer at 600.00 MHz (for 1H) and 6.22%;found:C,35.37;H,3.26;N,5.94%.1HNMR(DMSO-d , 6 150.86 MHz (for 13C). The obtained 1H and 13C NMR spectra ppm): δ 10.80 (s, N2–H, 1H), 8.48 (d, J = 5.5 Hz, C6–H, 2H), were calibrated against the residual signals (2.50 ppm for 1H 7.92 (t, J = 7.8 Hz, C4–H, 2H), 7.21 (d, J = 7.3 Hz, C3–H, 2H), NMR,39.5ppmfor13CNMR) oftheusedsolvent.42 Thesplit- 7.14 (d, J = 6.4 Hz, C5–H, 2H), 5.99 (d, J = 6.4 Hz, C13–H, ting of the 1H NMR signals is defined as s = singlet, d = C15–H,2H),5.90(d,J=6.4Hz,C12–H,C16–H,2H),2.44(sep, doublet, t = triplet, sep = septet, br = broad band, m = multi- J=7.2Hz,C17–H,1H),2.07(s,C20–H,3H),1.13(d,J=7.2Hz, plet. Mass spectrometry was performed on methanolic solu- C18–H, C19–H, 6H). 13C NMR (DMSO-d , ppm): δ 154.8 (C6), 6 tions using a LCQ Fleet Ion Trap spectrometer (Thermo 151.3 (C2), 140.6 (C4), 119.4 (C5), 113.3 (C3), 95.6 (C11), 91.4 Scientific;QualBrowsersoftware,version2.0.7)inthepositive (C14), 75.5 (C13,15), 75.3 (C12,16), 30.4 (C17), 21.9 (C18,19), electrospray ionization mode (ESI+). FTIR spectra were 17.6(C20). recorded bya Nexus 670 FT-IR spectrometer (Thermo Nicolet) [Os(η6-pcym)(dpa)Cl]BPh 4 (1-Cl′).Thecomplex1-Cl′waspre- usingtheATRtechnique(the400to4000cm −1region). pared using NaBPh instead of NH PF . Crystals of the 4 4 6 complex 1-Cl′·CH OH suitable for single-crystal X-ray analysis 3 werepreparedbyslowevaporation(ca.5days)ofamethanolic solutionofthecomplex1-Cl′. SinglecrystalX-rayanalysis [Os(η6-pcym)(dpa)(VP)]PF (1-VP). The complex 1-Cl X-ray data collection of the complex [Os(η6-pcym)(dpa)Cl] 6 (0.1 mmol) was dissolved in methanol (50 mL); 0.35 mmol of BPh ·CH OH (1-Cl′·CH OH) was performed on a D8 QUEST 4 3 3 AgVP was added, and the reaction mixture was stirred over- diffractometer (Bruker) using a PHOTON 100 CMOS detector nightunderaluminiumfoilatambienttemperature.Then,the and Mo–Kα radiation (λ = 0.71073 Å). The datacollection and white precipitate (AgCl) was filtered and NH PF (0.5 mmol) reduction were performed using the APEX3 software 4 6 was added to the obtained solution, which was concentrated package,43 and the structure was solved by a direct method to ca. 1 mL. An excess of diethyl ether was added, and the (SHELXS) and refined using the Bruker SHELXTL Software formed light yellow solid was collected by filtration, washed Package.44 The structural data were deposited in the with diethyl ether (3 × 2 mL) and dried under vacuum in a Cambridge Crystallographic Data Centre under the access desiccator.Theyieldwasca.55%(relativeto1-Cl).Anal.calcd number CCDC 1515628.† The graphics were drawn and for OsC H N O PF : C, 42.91; H, 4.89; N, 5.36%; found: C, additional structural calculations were performed using 28 38 3 2 6 42.53;H,5.01;N,5.05%.1HNMR(DMSO-d ,ppm):δ10.78(s, DIAMOND45 and Mercury46 software. The crystal data and 6 N2–H, 1H), 8.58 (d, J = 5.3 Hz, C6–H, 2H), 7.87 (t, J = 7.5 Hz, structurerefinementcanbefoundinTableS1.† Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online Paper DaltonTransactions 1HNMRstudiesofhydrolysis at an 80.0 μM concentration due to its lower solubility). After that, the solutions were removed, and the cells were washed Appropriate amounts (for 600 µL of 1 mM solutions) of the with drug-free medium and maintained in the drug-free complexes 1-VP and 1-Cl were dissolved in MeOD-d (60 µL); 4 then,eitherD O(540µL)orphosphatebufferinD O(540µL) mediumforthenext72h(recoverytime).Afterthetreatment, 2 2 was added. 1H NMR spectra were recorded for the fresh solu- the cell viability was assessed using the MTTassayand evalu- atedspectrophotometricallyat540 nmbyanInfinite 200PRO tions(t=0h)andafter0.5,1,2,4,6,24and48hofstanding at ambient temperature. The obtained 1H NMR spectra were microplate reader (TESCAB, Scholler Instruments LLC). The calibratedagainsttheresidualsignalofD O(4.85ppm).42For cytotoxicitydatawereobtainedfromthreeindependentexperi- 2 ments(eachconductedintriplicate)usingthecellsfromthree the solution of the complex 1-Cl in D O, 5 mol equiv. of KCl 2 consecutive passages. The IC values (µM) were calculated wasaddedafter48hofstandingatambienttemperature,and 50 the 1H NMR spectrum was recorded. For comparative pur- fromtheviabilitycurves,andtheresultsarepresentedasarith- poses, the 1H NMR spectra were also recorded for sodium meticmean±SD. valproate(NaVP)inbothmedia. Cellmorphologystudies Thecomplex1-OD (60µLofMeOD-d solutions;seeabove) 2 4 was mixed with 540 µL D O to afford three 1 mM solutions. A2780 cells were seeded in 6-well culture plates (5 × 105 cells 2 The pH of two of these solutions was adjusted by addition of perwell)andincubatedovernight (37°C,5%CO ,humidified 2 0.1%DClO and0.1%KODtovaluesof4and10,respectively, incubator). After that, the A2780 cells were treated for 24 h 4 andthe1HNMRspectrawererecorded(300K,JEOLJNM-ECA withtheIC concentrationsofcomplex1-VPandcisplatin(for 50 600II). comparative purposes), while the cells in one well were left untreated (negative control). After incubation for 24 h, the 1HNMRstudiesofinteractionswithGSH solutions were removed from the cells and the cells were washedwithdrug-free medium,in whichthecellsweremain- Appropriate amounts (for 600 µL of 1 mM solutions) of the tained for the next 72 h (recovery time). Then, the cells were complexes 1-VP and 1-Cl were dissolved in MeOD-d (60 µL), 4 stained by a standard hematoxylin/eosin staining procedure and a solution of GSH (2 mol equiv.) in either phosphate buffer in D O (540 µL; pH 7.2) or in a mixture of 0.1% KOD and observed under an NIB-100 inverted light microscope 2 and D O (540 µL; pH 7.2) was added. 1H NMR spectra were (Novel Optics Co., Ltd) equipped with an HDCE-50B digital 2 camera. recorded at various time-points (t = 0, 0.5, 1, 2, 4, 6, 24 and 48 h of standing at ambient temperature). The spectra were Cellularaccumulation referenced against the residual signal of D O found at 2 4.85 ppm.42 For comparative purposes, 1H NMR spectra were A2780cellswereseededin6-wellcultureplatesatadensityof also recorded for free GSH in both media. A similar experi- 1 × 106 cells per well and incubated overnight (37 °C and 5% ment was performed for a mixture of the complex 1-VP with CO in a humidified incubator). Then, the cells were treated 2 2 molar equiv. of GSH in 10% MeOH/90% H O with the for 24 h with the complexes 1-VP and 1-Cl in concentrations 2 additionof0.1%KOH(pH7.2)andmonitoredbyESI+/−mass corresponding to their IC values. Thereafter, cells were 50 spectrometry. washed with PBS (2 × 2 mL), harvested by trypsinization, and centrifuged. The supernatants were discarded and the pellets Cellcultureandinvitrocytotoxicitystudies were immediately digested in 500 µL of nitric acid (3 min, 150°C)usingamicrowavereactionsystem.Thesolutionswere The A2780 human ovarian carcinoma, MCF-7 human breast diluted with 4.5 mL of water, and the osmium content was adenocarcinoma, PANC-1 human pancreatic adenocarcinoma determined by ICP-MS (ICP-MS spectrometer 7700x, Agilent). andHT-29humancoloncarcinomacelllinesweresuppliedby Theobtainedvalueswerecorrectedforadsorptioneffects.The the European Collection of Cell Cultures (ECACC), while the experiments were performed in triplicate, and the results are humannon-canceroushepatocytes(Hep)werepurchasedfrom presentedasarithmeticmean±SD. BiopredicIntl.Thecellswereculturedasadherentmonolayers according to the supplier instructions in RPMI-1640 medium Flowcytometrystudies supplementedwith10%fetalcalfserum,1%2mMglutamine and 1% penicillin/streptomycin (37 °C and 5% CO , humidi- A2780cellswereseededin6-wellcultureplatesatadensityof 2 fiedincubator). 1 × 106 cells per well and incubated overnight (37 °C and 5% Following seeding in 96-well culture plates, the A2780, CO in a humidified incubator). The cultured cells were har- 2 MCF-7,PANC-1,HT-29andHepcellswereallowedtostabilize vestedbytrypsinization,washedwithPBS,centrifuged,re-sus- for 24 h. Then, the cancer cells and Hep cells were incubated pended in PBS, divided (individual samples counted ca. 3 × for 24 h (exposure time) with vehicle (DMF; 0.1% v/v), Triton 105cells)andstained,asspecifiedbelow.Thecomplexes1-VP X-100 (1%, v/v; positive control), cisplatin and the complexes and 1-Cl were applied at equipotent concentrations (equal to 1-VP and 1-Cl in concentrations of 0.1 to 100.0 μM (for the their IC values). The cells were analysed by a CytoFlex flow 50 A2780, MCF-7, PANC-1 and HT-29 cells) and 1.0 to 200.0 μM cytometer(BeckmanCoulter),andtheobtaineddatawereana- (for the Hep cells) (except forcisplatin, which wastested only lysed using CytExpert™ software (Beckman Coulter). In all DaltonTrans. Thisjournalis©TheRoyalSocietyofChemistry2018 .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online DaltonTransactions Paper cases, untreated stained A2780 cells were used as a negative All the measurements were carried out using a MiniScope control.Allexperimentswerecarriedoutintriplicate. MS5000 Magnettech X-band EPR spectrometer (Freiberg Cellcyclestudies.A2780cells(treatedfor2hand24h)were Instruments GmbH, Freiberg, Germany). The samples were stained for 30 min with propidium iodide (PI) supplemented measuredin50µLcapillariesimmediatelyafterpreparationat with RNase A (25 °C). After that, the cells were re-suspended ambienttemperature,usingthefollowingsettings:modulation and their DNA content was assessed using flow cytometry, amplitude 0.1 mT, modulation frequency 100 kHz, microwave detecting the emission at 617 nm after excitation at 535 nm. power10mW,sweepwidth20mT,scantime3min. A2780 cells treated with cisplatin (IC concentration) were 50 Statisticalanalysis usedforcomparativepurposes. Induction of apoptosis. The cells (treated for 2 h and 24 h) TheANOVAtestwasusedforstatisticalanalysis,withvaluesof were processed according to the instructions supplied in the p < 0.05 (*), 0.01 (**) and 0.005 (***) considered to be statisti- Annexin V-FITC Apoptosis Detection Kit (Enzo Life Sciences). cally significant. QC Expert 3.2 Statistical software (TriloByte Cells treated with the apoptosis-inductor staurosporine Ltd)wasusedtoperformtheanalysis. (1 μg mL −1) were used in the studyas a positive control. Flow cytometrywasperformedusingabluelaser(488nm)and525/ 40BPand610/20BPfilters. Conclusions ROS/superoxide assay. A2780 cells (24 h exposure) were stainedaccordingtothemanufacturer’sinstructionsusingthe In conclusion, we demonstrate for the first time that the orange/green fluorescent reagents supplied in the Total ROS/ invitrocytotoxicityofwidelystudiedhalf-sandwichOs(II)chlo- Superoxide detection kit (Enzo Life Sciences). Cells treated rido complexes can be enhanced by the insertion of a releas- with pyocyanin were used as the positive control. Cells were ablebioactiveligandfromthefamilyofHDACinhibitors,such analysedbyflowcytometry,detectingtheemissionsat525nm asvalproateusedinthiswork.Weusedaconvenientsynthesis (totaloxidativestress)and620nm(superoxide). starting from easily obtainable complex [Os(η6-pcym)(dpa)Cl] Mitochondrial membrane potential assay. A2780 cells were PF 6 (1-Cl), whose interaction with an excess of silver(I) valpro- treated for 24 h and then processed using a MITO-ID® ate (AgVP) provided the novel valproato complex [Os(η6-pcym) Membrane potential detection kit (emission maxima at 590 (dpa)(VP)]PF 6 (1-VP). The complex 1-VP displayed ca. 3-times and 525 nm for the orange and green dyes, respectively; Enzo higher in vitro cytotoxicity against the A2780 ovarian carci- Life Sciences). Stained cells treated with carbonyl cyanide noma cells than the 1-Cl analogue. Profound mechanistic 3-chlorophenylhydrazone (CCCP; 2 μM final concentration) investigationsindicatedthattheinsertionoftheHDACinhibi- were used asthe positive control. A2780 cellstreated with the tor valproate into the structure of the complex 1-VP also IC concentration of cisplatin were studied for comparative affectedtheonsetofapoptosis,whichwasslowerforthemore 50 purposes. cytotoxic complex 1-VP than for the complex 1-Cl. The herein obtained biological results for the complex 1-VP suggest that Cell-freehistonedeacetylaseassay this agent and its analogues are suitable for further investi- Histonedeacetylaseactivitywasevaluatedusingafluorometric gations, focusing on the intracellular fates of the released Histone Deacetylase Assay Kit (Sigma-Aldrich, Prague, Czech speciesanddual-actionstudies. Republic) processed according to the supplier’s instructions. The mixtures were incubated with the IC concentration of 50 Conflicts of interest the complex 1-VP (and VPA and cisplatin forcomparative pur- poses)for30minat30°C.Fluorescence(excitationat360nm) Therearenoconflictstodeclare. was recorded at 460 nm by an Infinite 200 PRO microplate reader. Acknowledgements EPRspin-trappingstudies The complexes 1-Cl and 1-VP were dissolved in DMF to afford We thank projects 17-08512Y, LO1305 and CZ.1.05/2.1.00/ 10 mM stock solutions; meanwhile, a 100 mM stock solution 19.0377 for financial support, Ms. Lucie Hanousková for her ofH 2 O 2 and60mMstocksolutionsofDEPMPOwereprepared helpwiththesyntheses,Ms.KateřinaKubešováforperforming in75mMPBS.Thesampleswerepreparedbymixing50μLof the in vitro cytotoxicity experiments, Dr. Bohuslav Drahoš for all three components (i.e., complex, H O and spin trap). The 2 2 performingESI+MSandRP-HPLC/ESI+MSexperiments,Mrs. blank samples contained eithercomplex and spin trap (50 μL Pavla Richterová for performing elemental analysis, and Dr. each,dilutedwith50μLof75mMPBS)orH O andspintrap 2 2 AlenaKlanicováforrecordingtheFTIRspectra. (50 μL each, diluted with 50 μL of DMF). To study possible quenching by relevant biomolecules, samples containing the complex 1-VP, H O and spin trap with the addition of 2 mol Notes and references 2 2 equiv.ofGSH(takenagainsttheOs(II)complex)wereprepared andmeasuredaswell. 1 L.Kelland,Nat.Rev.Cancer,2007,7,573. Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online Paper DaltonTransactions 2 (a)N.P.E.BarryandP.J.Sadler,Chem.Commun.,2013,49, 21 J. Patalenszki, L. Bíro, A. C. Benyei, T. Radosova Muchova, 5106; (b) M. Hanif, M. V. Babak and C. G. Hartinger, Drug J.KasparkovaandP.Buglyó,RSCAdv.,2015,5,8094. DiscoveryToday,2014,19,1640. 22 E. Păunescu, M. Soudani, C. M. Clavel and P. J. Dyson, 3 (a) Y. Fu, A. Habtemariam, A. M. Pizarro, S. H. van Rijt, J.Inorg.Biochem.,2017,175,198. D.J.Healey,P.A.Cooper,S.D.Shnyder,G.J.Clarksonand 23 W. Kandioller, E. Balsano, S. M. Meier, U. Jungwirth, P. J. Sadler, J. Med. Chem., 2010, 53, 8192; S.Göschl,A.Roller,M.A.Jakupec,W.Berger,B.K.Keppler (b) A. C. Matsheku, M. Y.-H. Chen, S. Jordaan, S. Prince, andC.G.Hartinger,Chem.Commun.,2013,49,3348. G. S. Smith and B. C. E. Makhubela, Appl. Organomet. 24 M. Groessl, O. Zava and P. J. Dyson, Metallomics, 2011, 3, Chem.,2017,31,e3852;(c)C.A.Riedl,L.S.Flocke,M.Hejl, 591. A.Roller,M.H.M.Klose,M.A.Jakupec,W.Kandiollerand 25 M. J. McKeage, S. J. Berners-Price, P. Galettis, R. J. Bowen, B.K.Keppler,Inorg.Chem.,2017,56,528. W. Brouwer, L. Ding, L. Zhuang and B. C. Baguley, Cancer 4 H. Kostrhunova, J. Florian, O. Novakova, A. F. A. Peacock, Chemother.Pharmacol.,2000,46,343. P.J.SadlerandV.Brabec,J.Med.Chem.,2008,51,3635. 26 C.Bachran,R.Hasikova,C.E.Leysath,I.Sastalla,Y.Zhang, 5 (a) I. Romero-Canelón, M. Mos and P. J. Sadler, J. Med. R. J. Fattah, S. Liu and S. H. Leppla, Cell Death Dis., 2014, Chem., 2015, 58, 7874; (b)J. M. Hearn,I.Romero-Canelón, 5,e1003. A. F. Munro, Y. Fu, A. M. Pizarro, M. J. Garnett, 27 M.L.CircuandT.Y.Aw,FreeRadicalsBiol.Med.,2010,48, U. McDermott,N. O. Carragherand P. J.Sadler,Proc.Natl. 749. Acad.Sci.U.S.A.,2015,112,E3800;(c)I.Romero-Canelón, 28 S. J. Dougan, A. Habtemariam, S. E. McHale, S. Parsons L.SalassaandP.J.Sadler,J.Med.Chem.,2013,56,1291. and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 6 R. Raveendran, J. P. Braude, E. Wexselblatt, 11628. V. Novohradsky, O. Stuchlikova, V. Brabec, V. Gandin and 29 V. Ribas, C. García-Ruiz and J. C. Fernández-Checa, Front. D.Gibson,Chem.Sci.,2016,7,2381. Pharmacol.,2014,5,151. 7 G. Agonigi, T. Riedel, M. P. Gay, L. Biancalana, E. Oñate, 30 S.Alas,C.P.NgandB.Bonavida,Clin.CancerRes.,2002,8, P. J. Dyson, G. Pampaloni, E. Păunescu, M. A. Esteruelas 836. andF.Marchetti,Organometallics,2016,35,1046. 31 J. M. Cross, T. R. Blower, N. Gallagher, J. H. Gill, 8 T. Rhodes and P. R. Twentyman, Br. J. Cancer, 1992, 65, K. L. Rockley and J. W. Walton, ChemPlusChem, 2016, 81, 684. 1276. 9 C. T. Chiu, Z. Wang, J. G. Hunsberger and D. M. Chuang, 32 C. Frejaville, H. Karoui, B. Tuccio, F. Le Moigne, Pharmacol.Rev.,2013,65,105. M. Culcasi, S. Pietri, R. Lauricella and P. Tordo, J. Med. 10 L. Činčárová, Z. Zdráhal and J. Fajkus, Expert Opin. Invest. Chem.,1995,38,258. Drugs,2013,22,1535. 33 S.StollandA.Schweiger,J.Magn.Reson.,2006,178,42. 11 (a) P. A. Marks, Expert Opin. Invest. Drugs, 2010, 19, 1049; 34 M. Culcasi, A. Rockenbauer, A. Mercier, J. L. Clement and (b) M. Gottlicher, S. Minucci, P. Zhu, O. H. Kramer, S.Pietri,FreeRadicalsBiol.Med.,2006,40,1524. A. Schimpf, S. Giavara, J. P. Sleeman, F. Lo Coco, C. Nervi, 35 F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, P.G.PelicciandT.Heinzel,EMBOJ.,2001,20,6969. 73. 12 J. M. Gichumbi, B. Omondi and H. B. Friedrich, 36 (a) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Eur.J.Inorg.Chem.,2017,2017,915. MatterMater.Phys.,1988,37,785;(b)A.D.Becke,J.Chem. 13 P. Štarha, Z. Trávníček, R. Křikavová and Z. Dvořák, Phys., 1993, 98, 1372; (c) A. D. Becke, J. Chem. Phys., 1993, Molecules,2016,21,1725. 98, 5648; (d) P. J. Stephens, F. J. Devlin, C. F. Chabalowski 14 D. M. Griffith, B. Duff, K. Y. Suponitsky, K. Kavanagh, andM.J.Frisch,J.Phys.Chem.,1994,98,11623. M. P. Morgan, D. Egan and C. J. Marmion, J. Inorg. 37 F.WeigendandR.Ahlrichs,Phys.Chem.Chem.Phys.,2005, Biochem.,2011,105,793. 7,3297. 15 N.Gulyaeva,A.Zaslavsky,P.Lechner,M.Chlenov,A.Chait 38 V.BaroneandM.Cossi,J.Phys.Chem.A,1998,102,1995. andB.Zaslavsky,Eur.J.Pharm.Sci.,2002,17,81. 39 (a) S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. 16 F. H. Allen, Acta Crystallogr., Sect. B: Struct. Sci., 2002, 58, Phys., 2010, 132, 154104; (b) S. Grimme, S. Ehrlich and 380. L.Goerigk,J.Comput.Chem.,2011,32,1456. 17 C. Romain, S. Gaillard, M. K. Elmkaddem, L. Toupet, 40 (a) V. Barone, in Recent Advances in Density Functional C. Fischmeister, C. M. Thomas and J. L. Renaud, Methods, Part 1, ed. D. P. Chong, World Scientific, Organometallics,2010,29,1992. Singapore, 1995, p. 287; (b) N. Rega, M. Cossi and 18 B. C. E. Makhubela, M. Meyer and G. S. Smith, V.Barone,J.Chem.Phys.,1996,105,11060. J.Organomet.Chem.,2014,772–773,229. 41 J.P.C.Coverdale,C.Sanchez-Cano,G.J.Clarkson,R.Soni, 19 J. Zajac, H. Kostrhunova, V. Novohradsky, O. Vrana, M.WillsandP.J.Sadler,Chem.–Eur.J.,2015,21,8043. R. Raveendran, D. Gibson, J. Kasparkova and V. Brabec, 42 H. E. Gottlieb, V. Kotlyar and A. Nudelman, J. Org. Chem., J.Inorg.Biochem.,2016,156,89. 1997,62,7512. 20 S.H.vanRijt,A.Mukherjee,A.M.PizarroandP.J.Sadler, 43 Bruker. Apex3, Bruker AXS Inc., Madison, Wisconsin, USA, J.Med.Chem.,2010,53,840. 2015. DaltonTrans. Thisjournalis©TheRoyalSocietyofChemistry2018 .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online DaltonTransactions Paper 44 G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 46 C. F. Macrae, I. J. Bruno, J. A. Chisholm, P. R. Edgington, 2015,71,3. P. McCabe, E. Pidcock, L. Rodriguez-Monge, R. Taylor, 45 K. Brandenburg, Diamond Version 4.0.3, Crystal Impact J. van de Streek and P. A. Wood, J. Appl. Crystallogr., 2008, GbR,Bonn,Germany,2015. 41,466. Thisjournalis©TheRoyalSocietyofChemistry2018 DaltonTrans. .30:94:11 8102/40/01 no nilreB teatisrevinU eierF yb dedaolnwoD .8102 hcraM 12 no dehsilbuP View Article Online