Antileukemic activity and cellular effects of rhodium(III) crown thiaether complexes.

Biometals(2011)24:645–661 DOI10.1007/s10534-011-9414-9 Antileukemic activity and cellular effects of rhodium(III) crown thiaether complexes Ruth Bieda • Igor Kitanovic • Hamed Alborzinia • Andreas Meyer • Ingo Ott • Stefan Wo¨lfl • William S. Sheldrick Received:31October2010/Accepted:12January2011/Publishedonline:28January2011 (cid:2)SpringerScience+BusinessMedia,LLC.2011 Abstract The cytostatic properties of novel rho- with LL = bpm, dpq, and 1,4-dithiane. The order of dium(III) thiacrown ether complexes [RhCl(LL)([9] activity (bpm[1,4-dithiane[dpq[bpy) contrasts aneS )]n?witheitheraromaticj2Nligands(n = 2)or tothatobservedforadhesivecancercells(bpm[bpy, 3 anionicchelateligands(n = 1)havebeeninvestigated 1,4-dithiane[dpq). Necrosis is insignificant in forthehumancancercelllinesHT-29andMCF-7and all cases. The percentage of Jurkat cells exhibiting for immortalized HEK-293 cells. Taken together apoptosisafter24or48 hincubationperiodsisdirectly with literature IC values for analogous complexes correlated to the percentage of cells exhibiting high 50 with polypyridyl ligands or 1,4-dithiane, the in vitro levels of reactive oxygen species. As established by assays indicate that dicationic complexes with soft onlinemonitoringwithasensorchipsystem,treatment j2N (imino) or j2S (thiaether) ligands exhibit signif- of MCF-7 cells with the bpm and 1,4-dithiane com- icantlyhigherantiproliferativeeffectsthanthosewith plexesleadstoasignificantandpermanentconcentra- hard j2N (amino) ligands. Dicationic complexes are tion-dependent decrease in oxygen consumption and moreactivethanmonocationiccomplexeswithsimilar cellularadhesion. ligands. Pronounced apoptosis-inducing properties towards Jurkat cells were established for complexes Keywords Rhodium (cid:2) Thiacrown ether (cid:2) Cytotoxicity (cid:2) Leukemia (cid:2) Apoptosis (cid:2) ROS Electronicsupplementarymaterial Theonlineversionof thisarticle(doi:10.1007/s10534-011-9414-9)contains supplementarymaterial,whichisavailabletoauthorizedusers. Introduction R.Bieda(cid:2)W.S.Sheldrick(&) Selective treatment is a prerequisite for all oncolog- Lehrstuhlfu¨rAnalytischeChemie,Ruhr-Universita¨t icaldrugsintumortherapy.Astherearecurrentlyno Bochum,44780Bochum,Germany e-mail:william.sheldrick@rub.de tumor therapeutic drugs known without serious side- effects, a high level of current interest is present for I.Kitanovic(cid:2)H.Alborzinia(cid:2)S.Wo¨lfl research on novel cell selective drugs. Metal com- Institutfu¨rPharmazieundMolekulareBiotechnologie, pounds can display an advantageous chemical func- Ruprecht-Karls-Universita¨tHeidelberg,ImNeuenheimer Feld364,69120Heidelberg,Germany tionality and are currently of increasing interest in medical chemistry, both as diagnostic agents and as A.Meyer(cid:2)I.Ott drugs(Jakupecetal.2008;OttandGust2007;Clarke Institutfu¨rPharmazeutischeChemie,Technische 2002). Only three metal based compounds, cisplatin, Universita¨tBraunschweig,Beethovenstraße55, 38106Braunschweig,Germany carboplatin and oxaliplatin, are currently in use as 123 646 Biometals(2011)24:645–661 antitumordrugs,andallofthesecontainplatinum(II) sequences,whereastheirstructuralanalogue[Rh([12] as the central metal atom. Cisplatin, the most aneS )(phi)]3?withtheHacceptorthiaethercoligand 4 successful metal-containing antitumor therapeutic [12]ane-S doesnot(ShieldsandBarton1995).Upon 4 agent (Rosenberg et al. 1969) is especially active irradiation at 311 nm, the phototoxic and photonuc- towards testicular and ovarian cancer. Nephro-, oto- lease agent [RhCl (dppz)(phen)]Cl exhibits toxicity 2 and neurotoxicity are unfortunate side-effects, which towards the tumor cell lines GN4, M109 and KB restrict its clinical utility to a certain extent. Repre- (Menonetal.2004). sentativeexamplesofnon-platinummetal-containing Recent work on the possible anticancer activity of compoundscurrentlyinclinicaltrialsare(ImH)[trans- otherrhodium(III)polypyridylcomplexeshasproduced {RhCl (DMSO)(Im)}] (NAMI-A) and (IndH)[trans- some highly promising results. Systematic studies on 4 {RhCl (Ind) }] (KP1019) (Jakupec et al. 2008). the cytotostatic properties of the compounds mer- 4 2 Whereas the high effectiveness of NAMI-A against [RhCl (DMSO)(pp)] and [(C Me )RhCl(pp)](CF SO ) 3 5 5 3 3 metastases is related to a combination of its anti- (pp = bpy,phen,dpq = dipyrido[3,2-d:20,30-f]quinoxa- angionesisandanti-invasivepropertiesagainsttumor line, dppz, dppn) have demonstrated that the in vitro cellsandvasculature(Morbidellietal.2003),KP1019 antiproliferative activities of these contrasting types of enteredclinicaltrialsforitsactivityagainstcolorectal rhodium(III)complexes towards thehumancancer cell cancer and their metastases (Hartinger et al. 2006; lines HT-29 (colon carcinoma) and MCF-7 (breast Rademaker-Lakhai et al. 2004). It has been reported, carcinoma) increase with increasing polypyridyl ligand that KP1019 can influence the mitochondrial mem- size (Harlos et al. 2008; Scharwitz et al. 2008). The brane stability and can activate caspase-3, which is meridional trichlorido complexes are extremely potent well-known for its induction of apoptosis by the and exhibit IC values in the range 0.051–0.095 lM 50 mitochondrial pathway (Kapitza et al. 2005). towards the MCF-7 and HT-29 cell lines for the larger Transition metal complexes containing polypyridyl polypyridyl ligands pp = dpq, dppz, dppn = benzo[i] (pp)ligandshaverecentlyreceivedconsiderableatten- dipyrido[3,2-a:20,30-c]phenazine (Dobroschke et al. tionaspotentialdiagnosticagentsowingtotheirability 2009). Significant levels of cell cytotoxicity have torecognizeandbindatspecificbasesequencesinDNA also been reported for other rhodium(III) compounds (Erkkila et al. 1999; Metcalfe and Thomas 2003). including mer-[RhCl (tpy)] (tpy = 2,20:6,200-terpyri- 3 Octahedral Ru(II) and Rh(III) compounds containing dine) (Pruchnik et al. 2002), fac-[RhCl ([9]aneNS )] 3 2 the smaller ligands 2,20-bipyridine (bpy) or 1,10- ([9]aneNS = 1-aza-4,7-dithiacyclononane) (Medvetz 2 phenanthroline (phen) have been shown to be groove et al. 2007) and polypyridyl complexes of the types binders or possible partial intercalators (Erkkila et al. [RhCl(pp)([9]aneS )]2?([9]aneS = 1,4,7-trithiacyclo- 3 3 1999;Zeglisetal.2007).Forlargerpolypyridylligands nonane) (Bieda et al. 2009a) and [RhCl(pp)(tpm)]2? like phi (9,10-diaminophenanthroline) (Sitlani et al. (tpm = tris(pyrazolyl)methane) (Bieda et al. 2009b) 1992) and dppz (dipyrido[2,3-a:20,30-c]phenazine) towards adhesive human cancer cell lines. We now (Frodletal.2002)strongintercalativebindinghasoften report an extension of our previous studies on rho- beenestablished. dium(III) crown thiaether compounds to aromatic [Rh(phi) (phen)]3?and[Rh(phi)(phen) ]3?bothdis- dinitrogen ligands other than pp (complexes 1–4) 2 2 play nuclease activity. Whereas [Rh(phi) (phen)]3? (Scheme 1)andtobidentateanionicligands(complexes 2 cleaves in a sequence neutral manner, [Rh(phi) 5 and 6) (Scheme 2). Particular emphasis has been (phen) ]3? cleaves 50-pyr-pyr-pur-30 sequences selec- placed on the biological impact of ([9]aneS )Rh(III) 2 3 tively possibly due to its steric and van der Waals complexesonnon-adhesiveleukemiacells. interactionswithinthemajorgroove[Sitlanietal.1992). These findings underline the possible influence of the ancillary ligands on both the DNA binding and the Results and discussion biologicalpropertiesofrhodiumpolypyridylcomplexes. The important role of ancillary ligands in enabling Synthesis and structure specificDNAsitebindingisalsodemonstratedbytheH donorligandsenand[12]aneN in[Rh(en) (phi)]3?and An in situ preparation procedure was established for 4 2 [Rh([12]aneN )(phi)]3?, which can recognize 50-GC-30 the novel ([9]aneS )RhIII complexes of the type 4 3 123 Biometals(2011)24:645–661 647 2+ Aromatic diamines of various lengths and shapes wereemployedasligandsin2–4toestablishwhether S the activity of these complexes is significantly S N different from that of compounds containing the Rh larger polypyridyl ligands dpqanddppz.Theligands S N phi, naphdiamine and dab exhibit the ability to bind Cl tometalatomseitherasdiiminesunderabstractionof twoprotonsorastertiarydiamines.Inthecaseofthe COOH H ([9]aneS )RhIII compounds 2–4, only 9,10-diamin- 3 N N H N 2 ophenanthrene lost two amino protons during the coordination process to produce the diimino phi N H N H 2 N complex2,asestablishedbyLSIMSanalysis.TheIR 4 spectrum of 2 lacks a strong d(NH ) band in the COOH 2 1 2 3 typical 1,550–1,640 cm-1 range for amino com- plexes and contains only the characteristic weak Scheme1 Structuresofthecomplexes1–4 intensity bands for coordinated phi, namely d(N–H) at 1,501 cm-1 and t(C=N) at 1,604 cm-1, near to or - + in this range (Madureira et al. 2000). For the S - N naphdiamine and dab complexes 3 and 4, LSIMS S dataclearlyindicateadiaminocoordination,whichis S L confirmed by the presence ofstrong t(NH )bands in 2 Rh S the range 3,250–3,000 cm-1 and strong d(NH ) 2 S L 5 bands between 1,610 and 1,550 cm-1 in both of Cl 6 their IR spectra, e.g. t(NH ) at 3,169, 3,134, 3,091 2 and3,045 cm-1andd(NH )at1,608and1,553 cm-1 2 Scheme2 Structuresofthecomplexes5and6 for the naphdiamine complex 3. The phenomenon of Ru-assisted proton abstraction and diimine coordina- [RhCl(LL)([9]aneS )]2?1–4(LL = bpy-4,40-(COOH) , tion has been previously reported for phi and dab 3 2 phi, naphdiamine, dab,) and [RhCl(phenylpy)([9] compounds (Madureira et al. 2000). aneS )]? 6 in accordance to that employed for the Pronounced antiproliferative activity (IC \12.8 3 50 previously published compounds [RhCl(LL)([9] lM) against MCF-7 and HT-29 cells has previously aneS )]2?7–14(LL = bpy,bpm =2,20-bipyrimidine, been established for ([9]aneS )RhIII complexes con- 3 3 phen, tap = pyrazino[2,3-f]quinoxaline, dpq, dppz, tainingthej2Sligand1,4-dithiane(Biedaetal.2010) pip = piperazine, 1,4- dithiane) (Bieda et al. 2009a; and the smallest pp ligand 2,20-bipyridine (Bieda 2010).TreatmentofRhCl (cid:2)3H Owiththeappropriate et al. 2009a). We have now employed the anionic 3 2 bidentate(pp)ligand ineitherCH OH,H O,CH Cl j2S ligand dmdtc- in ([RhCl(dmdtc)[9]aneS )]? 5 3 2 2 2 3 or CH OH/CH Cl (L) afforded the intermediate and the 2-phenylpyridine ligand phenylpy- in 3 2 2 complexes [RhCl (pp)(L)] in situ which were then [RhCl(phenylpy)([9]aneS )]Cl 6 to establish whether 3 3 employed without further separation or characteriza- reducing the overall complex charge to ?1 will lead tion for the preparation of the [9]aneS complexes. to an increase in activity due to a possible improve- 3 Compound[RhCl(dmdtc)([9]aneS )]?5wassynthes- ment in cellular uptake. 3 ised by addition of Na(dmdtc) (dmdtc = dim- ethyldithiocarbamate) to [RhCl (MeCN)([9]aneS )]? DNA interaction studies 2 3 (Bieda et al. 2010) in analogy to the literature procedureforsimilarcomplexeswiththedithiocarba- DNA interactions of the ([9]aneS )RhIII complexes 3 mate ligands R NCS - (R = CH CH , PhCH ) 1–6 were investigated together with those of the pre- 2 2 3 2 2 (Brandt and Sheldrick 1996). Complexes 1–6 were viously reported complexes [RhCl(pip)([9]aneS )]2? 3 characterizedby1HNMRandpositive-ionLSIMSand 13 and [RhCl(1,4-dithiane)([9]aneS )]2? 14 by CD 3 gavesatisfactorymicroanalyses. and UV/Vis spectroscopy. Whereas half-sandwich 123 648 Biometals(2011)24:645–661 organometalliccomplexeswithlabilechlorideligands 275 nm due to base stacking are observed. Changes typically bind to nucleobase nitrogen atoms (Annen induced by specific interactions between the biomol- et al. 2000; Sheldrick et al. 1993), those of the type ecule and the aromatic ligand of the transition metal [(g5-C Me )RhCl(pp)]? (pp = dpq, dppz) are strong complexes can induce CD bands between 280 and 5 5 intercalatorsintoDNA(HerebianandSheldrick2002; 400 nmasaresultofintercalation,surfaceorgroove Scharwitzetal.2008),andsuchabindinginteraction binding. Only very minor changes in the DNA has also been established for [RhCl(pp)([9]aneS )]2? conformation were established on addition of the 3 and[RhCl(pp)(tpm)]2?(pp = dpq,dppz)(Biedaetal. compounds 1, 2 or 6. Small differences are apparent 2009a, b). Following initial incubation periods of in Fig. S1 for the induced CD signal related to the 5 min, UV/Vis spectra of buffered solutions of com- base stacking of CT DNA in the presence of 5 or 14 plexes1–6(10 mMphosphatebuffer,pH = 7.2)with {r = [complex]/[DNA] = 0.2 for [DNA] = M (bp)} calf thymus DNA (CT DNA) at 25(cid:3)C and r = 0.1 in a 10 mM phosphate buffer. The molar ellipticity exhibited no further significant changes over 6 h, [h] decreased by 10% for 5 but increased by 5% for thereby indicating that achievement of equilibrium 14 at the maximum of the positive CD band at mustberapid.Effectivelynoabsorptionchangeswere 275 nm. observedforcomplexes[RhCl(LL)([9]aneS )]n?1–5, Smalldecreasesintheellipticityvalueforthebase 3 13 and 14 on mixing with CT DNA. Only minor stackingsignalwerealsoobservedforCTDNAinits changes (DA/A = 0.05 and 0.06) in the DNA mixtureswith2or4(Fig. 1).Instrikingcontrast,the max absorbance band between 250 and 260 nm were CD spectrum of compound 13 with CT DNA at observedfor6duringtheinitialmixingperiod. r = 0.2 displayed a pronounced decrease of 52% in DNA melting point changes in the presence of a the magnitude of the band related to base stacking. metalcompoundprovideastraightforwardmethodof Whereas the helix conformation is apparently little establishing the strength of DNA binding. The DT affectedbythisinteraction,thebasestackingsignalis m valuesforCTDNAinthepresenceofthecompounds reduced to nearly half its original value. This change [RhCl(LL)([9]aneS )]n?1–6were?2,?3,-2,-1,0 could result from covalent binding of the rhodium 3 and ?1(cid:3)C, respectively. The negative thermal dena- fragment to the double helix, for which additional turation temperature changes of -2 and -1(cid:3)C evidence is provided by the observed DT value of m invoked by the diamino complexes 3 and 4 are in -2(cid:3)C. Although this compound causes significant accordance with an absence of intercalation by their changesintheDNAconformation,itdoesnotexhibit aromaticligands.Incontrast, thecompounds 1and2 significant cytostatic activity (IC [100 lM) 50 with positive DT values of ?2 and ?3(cid:3)C may towards the adenomacarcinoma MCF-7 and HT-29 m exhibit groove binding or a limited degree of cells.Incontrast,whereasthediimino phicomplex2 intercalation. Following the loss of two amino causesonlyminorchangesforboththecharacteristic protons during the synthesis of compound 2, its DNA CD signals, it invokes relatively low IC 50 planar phi ligand can potentially interact with CT values of 19.20 (2.19) lM and 11.96 (1.93) lM in DNA in a manner similar to that of the diimino MCF-7 and HT-29. polypyridyl ligands of complexes 9–12. It is inter- esting to note that DT values of zero were recorded m for the j2S coordinated compounds 5 and 14. Both Cellular effects compounds have, therefore, no significant effect on the stability of the DNA. Antiproliferative activity towards adhesive MCF-7, In the presence of small molecules, characteristic HT-29 and HEK-293 cells changesinthecirculardicroism(CD)spectraofDNA in the range of 220–400 nm can provide a means of In vitro studies on the antiproliferative activities of monitoring possible conformational changes for the compounds of the type [RhCl(LL)([9]aneS )]n? 1–6 3 biopolymer(Brabecetal.1990;ErikssonandNorden were carried out with the human MCF-7 (breast 2001). CT DNA is present in the B DNA form for cancer) and HT-29 (colon cancer) cell lines and the which a negative CD band at 246 nm caused by the resulting IC values are listed in Table 1 together 50 helical conformation and a positive CD band at with those for the ligands LL in Table 2. Estimated 123 Biometals(2011)24:645–661 649 Fig.1 CD spectra of CT DNA and mixtures of [DNA]=0.2 for [DNA]=M (bp)} in a 10mM phosphate [RhCl(phi)([9]aneS )]Cl 2, [RhCl(dab)([9]aneS )]Cl 4 and buffer (pH=7.2) after an incubation period of 1h. Molar 3 2 3 2 [RhCl(pip)([9]aneS )]Cl 13 with CT DNA {r=[complex]/ ellipticities[h]aregivenintheunitsdegcm2dmol-1910-3 3 2 Table1 IC (lM)valuesforthecomplexes[RhCl(LL)([9]aneS )]n?1–14 50 3 Compound LL Counterions MCF-7IC HT-29IC HEK-293IC 50 50 50 1 bpy-4,40(COOH) Cl [100 [100 [100 2 2 2 Phi Cl 19.20(2.19) 11.96(1.93) 26.1(5.0) 2 3 Naphdiamine Cl 18.06(3.42) 50.7(5.8) [100 2 4 Dab Cl [100 [100 [100 2 5 Dmdtc CF SO 33.1(2.1) 40(27) 94.1(0.5) 3 3 6 Phenylpy Cl 57.3(9.0) 51.6(17.3) 54.1(7.2) 7a Bpy Cl 12.8(0.2) 4.4(0.1) [100 2 8a Bpm Cl 1.7(0.5) 1.9(0.1) 0.6(0.2) 2 9a Phen [PF ] 36.3(6.0) 72.2(8.0) [100 62 10a Tap Cl 11.5(0.6) 7.6(4.8) 3.3(1.3) 2 11a Dpq Cl 19.1(0.3) 20.9(2.8) 45.5(1.9) 2 12a Dppz Cl 4.7(0.5) 7.4(2.2) 8.9(1.6) 2 13b Pip Cl [100 [100 [100 2 14b Dithiane Cl 9.6(2.2) 8.1(3.7) 11.6(2.9) 2 Estimatedstandarddeviationsaregiveninparentheses a Biedaetal.(2009a) b Biedaetal.(2010) standard deviations are given in parentheses. To The expected trend of increasing antiproliferative determine a possible cell selectivity of facial Rh(III) activity with increasing length of the polypyridyl compounds towards cancer cell lines, these com- ligand (Scha¨fer et al. 2007) has been established for pounds were also tested towards immortalized HEK- the analogous tpm compounds (Bieda et al. 2009b) 293 (human embryonic kidney) cells (Table 1). againstthecancercelllinesMCF-7andHT-29andin 123 650 Biometals(2011)24:645–661 Table2 IC (lM)valuesforselectedligands established for complex 3 with its larger aromatic 50 surface area. As no activity was observed towards Ligand MCF-7IC HT-29IC HEK-293IC 50 50 50 HEK-293 cells, it is apparent that the cell lines bpy(COOH) [100 [100 ND exhibitincreasingsensitivitytowardsthiscomplexin 2 phi 19.5(1.5) 15.6(3.0) 29.2(9.6) the order MCF-7[HT-29 (cid:3) HEK-293, which may naphdiamine 43.9(29.0) 63.5(4.5) [100 berelatedtothecytotostaticpropertiesofthediamino dab 37.0(12.6) 58.8(16.6) 65.6(4.0) ligandonitsown.Fornaphdiamineitself,IC values 50 phenylpy [100 [100 [100 of 43.9(29.0) and 63.5(4.5) lM were observed Cisplatina 2.0(0.3) 7.0(2.0) 2.6(0.6) towards MCF-7 and HT-29 cells, respectively, but theIC valuewasgreaterthan100 lMforHEK-293 Estimatedstandarddeviationsaregiveninparentheses 50 cells. a Scharwitzetal.(2008) Apoptosis induction in the leukemia cell lines the case of the [9]aneS compounds for 9, 11 and 12 K562 and Jurkat 3 containing the larger polypyridyl ligands phen, dpq and dppz (Bieda et al. 2009a). This trend is also Weextendedthebiologicalstudieson([9]aneS )RhIII 3 followed by the IC values for the ([9]aneS )RhIII complexes to the non-adhesive leukaemia cell lines 50 3 compounds towards HEK-293 cells. A different K562andJurkat.Thecompounds7,8,10,11and14 mechanism of action was postulated for compounds were selected on the basis of their relatively high containing the nitrogen-rich bpm and tap ligands on antiproliferative activities (Table 1) towards the thebasisoftheirhighantiproliferativeactivity(Bieda adhesive cell lines MCF7 and HT-29 and their et al. 2010). The results presented in Table 1 allow differingtypesofbidentateligands.Apoptosisinduc- two structure–activity relationships tobeestablished. tion was tested for K562 (human chronic myeloge- Firstly,complexescontainingdiaminoligandsare,in nous leukemia) and Jurkat cells (human T-cell general, less active than j2N (imino) coordinated leukemia) by an Annexin V/propidium iodide assay complexes.Secondly,asexemplifiedbycomplex6in and PARP cleavage. The biological impact of the comparison to its j2N (imino) analogue 7, overall compounds was followed by measuring the intracel- charge reduction from ?2 to ?1 appears to lead to a lular level of reactive oxygen species (ROS) and by loss of activity. It is interesting to note, in this online monitoring of cellular metabolic processes. respect, that following deprotonation of its carboxyl- In apoptotic cells, phosphatidylserine is translo- ate functions, the inactive complex 1 will exhibit catedfromtheinnertotheouterleafletoftheplasma overall neutrality at biologically relevant pH values. membrane,therebyexposingittotheexternalcellular Howeverareducedlevelofcellularuptakecouldalso environment(vanEngelandetal.1998),whereitcan beresponsibleforitslackofantiproliferativeactivity. be detected by its binding to the intracellular protein IC values[100 lM towards all three cell lines Annexin V (Andree et al. 1990), labeled with a 50 wereestablishedforthediaminocomplexes4and13. fluorophore (Koopman et al. 1994). An Annexin V In comparison, the diimino phi complex 2 exhibits assayisabletodistinguishapoptoticcellsfromviable valuesof19.20(2.19)(MCF-7),11.96(1.93)(HT-29) or necrotic cells on the basis of the resulting fluores- and26.1(5.0) lM(HEK-293),thatarecomparableto cence of the former cells. A simultaneous dead cell those of the dpq complex 11. When compared to the stainemploysthefluorochromepropidiumiodide(PI), likewise very similar values of 19.5 (1.5) and 15.6 that can only penetrate dead or dying cells that have (3.0) lM for the ligand itself towards MCF-7 and lost or are losing membrane integrity. PI intercalates HT-29cells,itcanbepostulatedthattheactivityof2 intoDNAtoproduceahighlyfluorescentadduct,that may primarily be due to the spatial and stacking enables a precise evaluation of cellular DNA content properties of the phi ligand on its own. Whereas the by flow cytometric analysis (Lecoeur 2002; Riccardi dab compound 4 exhibits no significant antiprolifer- and Nicoletti 2006). PI positivity is an indicator of ative activity towards either MCF-7 or HT-29 cells eitherlateapoptosisornecrosis. (IC [100), IC values of 18.06 (3.42) lM and Annexin V/PI assays were first performed for 50 50 50.7 (5.8) lM towards the cancer cell lines were K562 cells with complexes 7, 10 and 11. After 48 h 123 Biometals(2011)24:645–661 651 incubation periods with concentrations of 10, 20 and 40 lM, the cells were, however, only affected in a minor fashion (data not shown). The percentage of viable cells was higher than 80% in all cases. Mock (solvent of the highest concentration) and 5 lM CMPT (camptothecin) samples were employed as controls.AssayswerethencarriedoutforJurkatcells after 24 and 48 h incubation periods with the same complexes. No significant apoptosis was determined after 24 h, which indicates that reduced cell viability is a time-dependent process (data not shown). Fig.2 Annexin V/PI assay towards Jurkat cells illustrating Figure 2 depicts the state of the Jurkat cells for each thenumberofcells/mlfollowingtheir48htreatmentwiththe of the three compounds in correlation to the number compounds 7, 10 and 11 at various concentrations. N gives of cells/ml present after the 48 h incubation periods. the number of cells/ml. Mock (solvent of the highest Allthreecompoundsreducedthetotalcellnumberin concentration)and5lMcamptothercin(CMPT)sampleswere employedascontrols a concentration-dependent manner. In comparison to the control value of nearly 6 9 105 cells/ml, com- pound 7 containing the smaller bpy ligand lowered cells correlating to increasing relative amounts of thenumbersofcellsonlygraduallyindependenceon early and late apoptotic cells were more apparent for its concentration to a final cell number of nearly compound 10 and particularly for 11 after the longer 4 9 105 cells/ml for a 40 lM solution. After treat- incubationperiod.Compound11inducesapoptosisin ment with a 10 lM solution of the tap complex 10, a clearly concentration-dependent manner. After a the viable cell number was lowered to about 5.3 9 48 h incubation with a 10 lM solution of 11, 87% 105 cells/ml and this value was further reduced to viable Jurkat cells were established, for a 20 lM about 3.5 9 105 cells/ml at a concentration of concentration 74% and for a 40 lM concentration 40 lM. The most pronounced cell number decrease only 63%. In conclusion, the data of the Annexin was determined in the case of complex 11 with the V/PI assay for the compounds 7, 10 and 11 towards largest polypyridyl ligand. After treatment with a the Jurkat cells demonstrate that with increasing size 10 lM concentration of the dpq complex about of the polypyridyl ligand (bpy\tap\dpq) increas- 2.9 9 105 cells/ml were still viable. Only 2 9 105 ing apoptosis induction is invoked within a 48 h cells/ml were viable in the case of a 20 lM solution incubation period, whereas after 24 h no significant and this number fell to 1.4 9 105 cells/ml for the apoptosis induction could be established. This order 40 lM solution. The extent of apoptosis induction ofactivitycontraststothatobservedfortheadhesive was not as high as observed for the CMPT control, MCF-7 and HT-29 cells (dpq\bpy\tap) and but the reduction in the number of cells/ml to about suggests that cell-type differences in the mechanism only a quarter of the control value is indicative of of action may play a role. pronounced anti-proliferative properties of the dpq- These findingspromptedustostudytheimpact of containing ([9]aneS )RhIII complex towards Jurkat two further rhodium(III) complexes 8 and 14 on 3 cells. It is interesting to note that complex 11 has a Jurkat leukemia cells. The former bpm complex markedly lower antiproliferative effect towards the exhibits the lowest IC values of all previously 50 adhesivecelllinesMCF-7andHT-29thancomplexes tested ([9]aneS )Rh(III)complexes (Table 1)and the 3 7 and 10 (Table 1). The findings of Fig. 2 therefore latter compound contains an aliphatic j2S chelate indicate an increased sensitivity of Jurkat cells ligandratherthanthearomaticj2N(imino)ligandsof towards 11 in comparison to 7 and 10. 7, 8, 10 and 11. Effectively no cell death inducing properties were In the case of complex 8 a high antiproliferative established for compound 7 towards the Jurkat cells. activity was also established for the non-adhesive The percentage of viable cells decreased only mar- Jurkatcells(Fig. 3).Nonecroticcellswereobserved. ginallyfrom96to93%withincreasingconcentration Evenatthelowestconcentrationof1.25 lM,thecell (10–40 lM) of 7. Decreasing percentages of viable number was reduced after 48 h to about a half in 123 652 Biometals(2011)24:645–661 PARP cleavage assay For confirmation of caspase activity and apoptosis induction,WesternBlot analyses were performed for possible PARP cleavage products. The DNA repair enzyme PARP [poly (ADP-Ribose) polymerase] is expressed more in apoptotic cells due to DNA fragmentation and is cleaved by caspases, especially Fig.3 AnnexinV/PIassaytowardsJurkatcellsillustratingthe by caspase-3 in vivo (Nicholson et al. 1995). The number of cells/ml following their 48h treatment with the inhibition and cleavage of PARP protein can there- compound [RhCl(bpm)([9]aneS )]2? 8 at various concentra- 3 fore be considered as being molecular markers for tions.Ngivesthenumberofcells/ml caspase-mediated apoptosis. An antibody against comparison to the mock control sample. At 20 lM PARP can detect intact PARP (116 kDa) and its only 1.0 9 105 cells/ml were present in comparison cleavage products (89 and 24 kDa). Western Blot to 3.6 9 105 cells/ml for the control. Following analysis for anti-b-actin obtained from the same treatment with a 5 lM solution nearly all cells were membranewasusedasaloadingcontroltoverifythe in an early or late apoptotic state. On increasing the total protein content. The Western Blot analyses of concentrationupto10and20 lM,thepercentagesof possible cleavage products of the PARP protein in cells undergoing early or late apoptosis were compa- Jurkat cells after a 24 h treatment at the indicated rable to those found for cells treated with CMPT at concentrations of the substances 11 and 7 are 10 lM. depicted in Fig. 5. This pair was chosen to enable a Figure 4 illustrates the numbers of Jurkat cells/ml comparison of possible caspase activity for com- followingtreatmentwiththe1,4-dithianecomplex14 plexes with markedly different activities towards the atvariousconcentrationsfor48 h.Thoughlowerthan leukemia cells. for the bpm complex8, amuchstronger antiprolifer- A 10 lM solution of complex 11 induced a ative effect was apparent in comparison to the poly- significant degree of apoptosis in Jurkat cells and pyridylcomplexes7and11.Whilethepercentageof activation of mitrochondia-dependent caspase cas- viable cells was similar to the percentage of late and cadesledtoPARPcleavagetothe89 kDaproductby earlyapoptoticcellsfora10 lMsolution,novitalcells caspase-3. The 89 kDa fragment was much less couldbedetectedata20 lMconcentration.Although intensive for a 20 lM solution and could not be its antiproliferative properties are less pronounced at clearlyvisualizedataconcentrationof40 lMdespite lower concentrations (\5 lM), the dithiane complex comparable loading. As this fragment just represents 14 exhibits similar apoptosis-inducing properties to aninitialstepinPARPcleavageanddegradationand the bpm complex 8 at higher concentrations. As for the process of caspase-mediated apoptosis is accel- the dpq complex 11, the Jurkat cells appear to be erated and/or enhanced at higher concentrations, it significantly more sensitive towards 14 than are the can be assumed that further PARP degradation must adhesive MCF-7 orHT-29cells. Fig.5 WesternBlot(WB)analysesofJurkatcellsafter24h Fig.4 AnnexinV/PIassaytowardsJurkatcellsillustratingthe treatment.aanti-PARP(top)with116kDaPARPand89kDa number of cells/ml following their 48h treatment with the cleavedPARPandbWesternBlot:anti-b-actin(42kDa)used compound [RhCl(dithiane)([9]aneS )]2? 14 at various asadetectorfortheappliedamountofproteinineachcase.NT 3 concentrations nottreated 123 Biometals(2011)24:645–661 653 have taken place for the 20 and 40 lM solutions of pronounced antileukemic activity were determined 11. For the effectively inactive compound 7 (see for Jurkatcells after 24and 48 h treatments. Fig. 2), no significant PARP cleavage product of Significantly enhanced levels of intracellular ROS 89 kDa could be determined even at the higher were observed after 24 h treatments with each of the concentrationsof20and40 lM.5 lMCamptothecin was used as a positive probe for caspase-induced apoptosis and PARP cleavage. Measurements of reactive oxygen species (ROS) An increase of intracellular reactive oxygen species (ROS) is correlated to a deregulated mitochondrial activity,whichisanadditionalindicationofapoptosis induction and may reflect overall stress. The occur- rence of ROS was measured for Jurkat cells by flow cytometryanalysisandisshowninFigs. 6,7and8as high and low ROS levels. After given incubation periods, the detection marker dihydroethidium bro- mide (DHE) was applied and the levels of high and low ROS were measured by FACS (fluorescence- activated cell sorting). The effects of the less active complex7,andthethreecompounds8,11and14with Fig.7 GenerationofreactiveoxygenspeciesinJurkatcellsin the presence of [RhCl(bpm)([9]aneS )]2? 8 after incubation 3 timesofa24handb48h Fig.6 Percentages of cells exhibiting high ROS levels in Jurkatcellsinthepresenceof[RhCl(bpy)([9]aneS )]2?7and Fig.8 Percentages of cells exhibiting high ROS levels in 3 [RhCl(dpq)([9]aneS3)]2? 11 after incubation times of a 24h Jurkatcellsinthepresenceof[RhCl(1,4-dithiane)([9]aneS )]2? 3 andb48h 14afteranincubationtimeofa24handb48h 123 654 Biometals(2011)24:645–661 complexes at a 40 lM concentration. After a 48 h does not contain an aromatic diimino donor ligand, incubation, the percentages of cells exhibiting high whose abilitiy to invoke ROS through either irradi- ROS levels differed more strikingly than for the ation or through reactions with cellular reductants is shorter time period (Fig. 6). Intracellular ROS were well-documented (Levina et al. 2009; Wang et al. generated at significantly higher levels at all concen- 2010). trations of 11. For compound 7, the increases in the percentagescellsexhibitinghighROS(17–26%)now Cellular metabolism correlated with increasing concentration. However ROS levels did not significantly differ from those A cell-based sensor chip system was employed to recorded after only 24 h. 11 is clearly more active monitor the influence of complexes 8 and 14 on the than 7 after 48 h. 20 and 40 lM solutions of 11 metabolism of MCF-7 cells. This system allows the induced high ROS levels for 70 and 74%ofthe cells online evaluation of the impedance of the cell layer, (Fig. 6b), whereas CMPT only induced high ROS oxygen consumption and the extracellular acidificat- levels in 52% of cells at a concentration of 20 lM. ion rate (glycolysis) of living cells over an extended It is instructive to compare Fig. 6 with the results time span. Its silicon chip includes interdigitated obtained for the potent bpm compound 8 after 24 electrodestructuresformeasuringthecellularimped- and 48 h (Fig. 7). This also invoked increasing ance (Ehret et al. 1998), miniature Clarke-type percentages of cells with high ROS levels with oxygenelectrodesformonitoringthecellularoxygen increasing compound concentration. Whereas at a uptake (Wolf et al. 1998) and ion-sensitive field 1.25 lM concentration, 27% of the Jurkat cells effect transistors to record extracellular pH changes already exhibited high ROS levels, this number rose (Lehmann et al. 2000). to 83% for a 20 lM solution of 8 following a 48 h The impedance of the MCF-7 cell layers fell incubation period (Fig. 7b). However, in striking rapidlyalmostimmediatelyaftertreatmentbeginwith contrast to the less active dpq complex, the percent- 50 lM solutions of 8 and 14 and reached a steady agesofcellsexhibitinghighROSlevelswerealready percentagelevelofabout25%ofitsinitialvalueafter very high after 24 h and comparable to those 20 h in both cases (Fig. 9). This decrease indicated observed for the longer incubation period. either morphological changes and/or changes in the Figure 8 displays the concentration-dependent status of cellular adhesion properties, including cell– percentages of Jurkat cells with high ROS levels cell and cell–matrix contacts, corresponding to the after 24 and 48 h treatments with 14. Incubation of the leukemia cells with a 10 lM solution of 14 induced threefold more cells with high ROS than CMPT and at a concentration of 20 lM of 14 more than 90% of the cells exhibited high ROS levels. As alsoestablishedforthepotentcomplex8,noeffective increase in the percentages of cells exhibiting high ROSlevelswasobservedinthetimeperiod24–48 h. In summary, a clear correlation between the percentage of cells exhibiting high ROS levels and the percentage of apoptotic cells was apparent after 48 h treatments with the potent complexes 8, 11 and 14. This finding is in accordance with studies that have indicated that higher levels of reactive oxygen species have a destructive effect on DNA and some proteins and that the associated oxidative stress can triggerapoptosis(Simonetal.2000;Sandstrometal. Fig.9 Standardcellimpedance(%)forMCF-7cellsovera38 hexposureperiodwith5and50lMsolutionsofcompounds5 1994). It is particularly interesting, that the aliphatic and 8. The end of treatment is indicated by a vertical line. 1,4-dithiane complex 14 induced high ROS concen- Measurements were continued for an additional 9h after trations similar to those of complex 8, although it removalofthesubstances.RMrunningmedium 123 Biometals(2011)24:645–661 655 Fig.11 Standard extracellular acidification rates (%) for Fig.10 Standard respiration rates (%) for MCF-7 cells over MCF-7 cells over a 38h exposure period with 5 and 50lM a 38h exposure period with 5 and 50lM solutions of solutions of compounds 5 and 8. The end of treatment is compounds 5 and 8. The end of treatment is indicated by a indicatedbyaverticalline.Measurementswerecontinuedfor vertical line. Measurements were continued for an additional anadditional9hafterremovalofthesubstances.RMrunning 9hafterremovalofthesubstances.RMrunningmedium medium Extracellular acidification is closely linked to the inductionofcelldeath.Theimpactof5 lMsolutions activity of glycolysis. This parameter is chiefly first became apparent after 16 h. Aless rapiddecline influenced by lactic acid production, which is a in the impedance of the cellular layer was then waste product of anaerobic metabolism. 50 lM observedbutthiscontinuedduringthefinaldrugfree solutions of 8 and 14 both invoked significant period (43–52 h), which indicated the infliction of decreases in the extracellular acidification rate for permanent cellular damage at concentration levels MCF-7 cells within the first hours of treatment close tothe IC values of8 and 14 (Figs. 3, 4). (Fig. 11), which indicates a lower cellular activity in 50 Oxygen consumption is generally indicative of comparison to nontreated cells. After reaching enhanced or decreased mitochondrial activity (respi- minimum values of about 80–88% following treat- ration).Otheroxygenconsumingprocessesaremuch ment for 20–30 h, the acidification rates increased less efficient and thus unlikely to contribute signif- slowly for cells treated with 5 lM solutions of the icantlytothissignal.FromFig. 10,itisapparentthat complexes during the final hours of treatment. the oxygen consumption of the MCF-7 cells was Interestingly, an increase was also observed for the dramatically affected during the first hours of treat- cellstreatedwiththe50 lMsolutionof8duringthis ment with 50 lM solutions of 8 and 14 and fell to a period, although their acidification rate had fallen to relative level of about only 10–15% after 16 and about only 35% of the original value after a 20 h 38 h, respectively. No recovery was observed in the exposure. All in all, the 38 h treatments had only a final drug free phase. Significant decreases in the minor long-term effect on the extracellular acidifi- respiration rates for the MCF-7 cells first became cation rate even at the higher concentration. apparent during the second day of treatment (28–43 h) with the 5 lM solutions. The relative valuesonlyfelltoabout70and80%forcellstreated Conclusions with 8 and 14, respectively, but no recovery was observedduringthedrugfreeperiod.Thispermanent The complexes [RhCl(LL)([9]aneS )]2? with LL = 3 reduction in cell respiration correlates well with the bpm (8) and 1,4-dithiane (14) are potent antiprolif- likewise dose-dependent increase in intracellular erativeagentstowardsbothadherentcancercellsand ROS levels in Jurkat cells following their treatment non-adherent leukemia cells. In contrast, a total lack with the complexes (Figs. 7, 8). of activity was established for the complexes 4 and 123 656 Biometals(2011)24:645–661 13 with the diamino chelate ligands 1,2-diaminoben- Materials and methods zene (dab) and piperazine (pip), which suggests that the presence of bidentate ligands with soft donor Chemicals atoms may well be a prerequisite for high activity in this type of complex. Reducing the overall charge RhCl (cid:2)3H OwaspurchasedfromABCR,1,4,7-trithia- 3 2 from?2to?1alsoledtoasignificantlossofactivity cyclononane, 9,10-diaminophenanthrene, sodium for the complexes 5 and 6. dimethyldithiocarbamate, 2-phenylpyridine and CT Complexes8and14induceapoptosisinJurkatcells DNAfromSigma–Aldrich,4,40-dicarboxy-2,20-bipyr- but cause negligible necrosis. A high percentage of idine from TCI, 2,3-diaminonaphthalene from Acros Jurkat cells exhibit high ROS levels after 24 h Organics, 1,2-phenylenediamine from Merck and treatments with these complexes at concentrations in crystal violet from Roth. Concentrations of CT DNA the 5–20 lM range and these percentages remain were determined spectrophotometrically using the effectivelyconstantoverthenext24 h.AnnexinV/PI molar extinction coefficient e = 13,200 M-1 cm-1 260 assays indicate that the apoptosis levels are directly (Marmur1961).Solventsforsynthesiswereanalytical correlated to the cellular ROS concentrations after reagentsgrade(JTBaker)andweredriedanddistilled 48 h, which suggests that the associated oxidative beforeuse.Deuteratedsolventsfor1HNMRmeasure- stress must play a central role in the mechanism of mentswereobtainedfromDeuteroGmbH. actionofthethiaethercomplexes.Cellularmetabolism studies demonstrated that the apoptosis induction is Instrumentation time-dependent at concentrations close to the com- pounds’ IC values. Cell death induction was first 50 UV/Vis spectra were recorded on an Analytik Jena apparent after about 16 h but then increased steadily SPECORD200,IRspectraonaNicoletImpact4000 over the following 24 h and during the drug free spectrometer. A Jasco J-715 instrument was used to period.Asimilartime-dependentpermanentdecrease measure CD spectra for complex/CT DNA mixtures. wasalsoestablishedforthecellularrespirationrate. Liquidsecondaryionmasspectrometrydata(LSIMS) Our present findings for complex 14 demonstrate were registered on a Fisons VG Autospec employing thatthepresenceofanaromaticdiiminochelateligand acesiumiongun(17 kV),with3-nitrobenzylalcohol is not essential for high antiproliferative activity but astheliquidmatrix.ABrukerDRX400spectrometer rather the presence of five soft donor atoms (i.e., 5 was used for the 1H NMR spectroscopic character- thiaetherSatomsor3thiaetherSatomsplus2iminoN ization of the new compounds, with chemical shifts atoms) in the octahedral coordination sphere of the reported as d values relative to the residual signal of rhodium(III) atom. The similarity in the ROS levels the deuterated solvent. The splitting of proton invoked by the complexes 8 and 14 in Jurkat cells resonances was defined as s = singlet, d = soublet, despite their very different bidentate ligands sug- m = multiplet. Elemental analyses were performed geststhatspecificnon-covalentinteractionsinvolving onaVarioELofElementarAnalysensystemeGmbH. the[9]aneS coligandand/orkineticfactorsmayplay 3 Compounds 7–14 were prepared in accordance with an important role in the mode of action of these literature procedures (Bieda et al. 2009a, 2010). rhodium(III)thiaethercomplexes.Duetotherestricted abilityofitsadditionalthiaetherSatomstoneutralize the high positive charge of the central rhodium(III) [RhCl(bpy-4,40-(COOH) )([9]aneS )]Cl (cid:2)H O 1 2 3 2 2 atoms through r donation (Brandt and Sheldrick 1996),asignificantlyslowerchloridesubstitutionrate 92.7 mg (0.38 mmol) 2,2-bpy-4,40-(COOH) were 2 will be expected for 14 in comparison to both the suspended in 20 ml ethanol and 100.2 mg j2N(imino)complex8and,inparticular,theinactive (0.38 mmol) RhCl (cid:2)3H O, dissolved in 20 ml etha- 3 2 piperazinecomplex13withitsmoreeffectiveaminoN nol, were added and the colourless suspension was r donor atoms. The slower decrease in MCF-7 cell refluxed at 100(cid:3)C for 2 h. Addition of 68.2 mg adhesion and respiration caused by 14 versus 8 (0.38 mmol)[9]aneS totheresultingorangesolution 3 (Figs. 9,10)may,therefore,reflectitsslowerreaction followedbyrefluxingat100(cid:3)Cforafurther4 hledto ratewithpossiblecellulartargetmolecules. formationofayellowsuspension.Afterfiltration,the 123 Biometals(2011)24:645–661 657 solution was reduced in volume to 5 ml and addition 5.27C27.0H4.53S18.02,found(%):N5.25C27.3H of diethylether led to precipitation of the product 4.0 S 18.5; LSIMS m/z (%) = 461.0 (15) [M-Cl]?, which was washed with EtOH and dried in vacuo. 425.0 (40) [M-Cl-HCl]?, 389.1 (93) [M-Cl-2HCl]?, Yield: 198.9 mg (80%), C H N O Cl S Rh(cid:2)H O 281.1(23)[M-Cl-2HCl-dab]?,329.1(17)[M-Cl-2HCl- 18 20 2 4 3 3 2 (651.85 g/mol) calcd. (%): N 4.30 C 33.17 H 3.4 S (CH CH S)]?,1H-NMR(DMSO-d ,400 MHz,25(cid:3)C): 2 2 6 14.76, found (%): N 4.1 C33.1 H 3.4 S 14.5; LSIMS d = 3.13–3.85(m,12H,CH [9]aneS ),7.17–7.75(m, 2 3 m/z (%) = 561.9 (87) [M-Cl-HCl]?, 526.9 (28) [M- 4H,dab)ppm. Cl-2HCl]?, 497.9 (43) [M-Cl-HCl-CH CH ]?, 465.9 2 2 (15)[M-Cl-HCl-SCH CH ]?,437.9(23)[M-Cl-HCl- 2 2 [RhCl(dmdtc)([9]aneS )](CF SO )(cid:2)2H O 5 S(CH CH ) ]?; 1H-NMR (MeOH-d4, 400 MHz, 3 3 3 2 2 2 2 25(cid:3)C): d = 3.05–3.89 (m, 12 H, CH [9]aneS ), 2 3 100 mg (0.18 mmol) of [RhCl (MeCN)([9]aneS )] 8.26 (d, 2H, H3/H8), 9.25 (dd, 2H, H2/H9) 9.37 2 3 (CF SO )(cid:2)1.5 H O [30] were dissolved in 15 ml (s, 2H, H5/H6) ppm. 3 3 2 MeOH and 33.2 mg of Na[S CNMe ] (0.18 mmol), 2 2 dissolved in 2 ml MeOH, were added. The reaction [RhCl(phi)([9]aneS )]Cl (cid:2)1.5H O 2 3 2 2 mixture was refluxed for 3 h. Following filtration of the yellow precipitate and volume reduction of the Preparation as for [RhCl(bpy)([9]aneS )]Cl [28] 3 2 remainingorangesolutionto5 ml,additionofdiethyl with 77.1 mg of phi (0.38 mmol), suspended in ether led to precipitation of the product, which was 10 ml CH Cl /EtOH/CH Cl (4/2/1). Yield: 170 mg 2 2 3 washed with MeOH and dried in vacuo. Yield: (72%); C H Cl N S Rh (cid:2)1.5H O (622.89 g/mol) 20 22 3 2 3 1 2 66.7 mg (61%), C H ClNO S F Rh(cid:2)2H O (622.85 calcd. (%): N 4.5 C 38.56 H 4.05 S 15.4, found 10 18 3 6 3 2 g/mol) calcd. (%): N 2.24 C 19.25 H 3.55 S 30.83, (%) N 3.7 C 38.8 H 4.0 S 14.9; LSIMS m/z found (%): N 1.8 C 18.8 H 3.1 S 30.4; LSIMS m/z (%) = 595.3 (14) [M-H]?, 524.3 (15) [M-Cl-HCl]?, (%) = 437.8 (100) [M-OTf]?, 402.8 (44) [M-HOTf- 443.3 (100) [M-S (CH CH ) ]?, 414.2 (56) [M- 3 2 2 2 Cl]? 374.8 (13) [M-HOTf-Cl-CH CH ]? 345.8 (4) [9]aneS -H]?, 1H-NMR (CD Cl , 400 MHz, 25(cid:3)C): 2 2 3 2 2 [M-HOTf-Cl-(CH CH ) ]?, 314.8 (13) [M-HOTf- d = 2.90–4.01 (m, 12H, CH [9]aneS ), 7.49 (t, 2H, 2 2 2 2 3 Cl–S(CH CH ) ]?, 1H-NMR (CD CN, 400 MHz, H3/H8),7.74(t,2H, H4/H7), 8.05–8.20(2d,4H,H2/ 2 2 2 3 25(cid:3)C): d = 3.19 (s, 6H, CH carbamate), 3.16–3.22 H9, H5/H6) ppm. 3 (m, 8H, CH [9]aneS ), 3.34–3.46 (m, 4H, CH 2 3 2 [9]aneS ) ppm. [RhCl(naphdiamine)([9]aneS )]Cl (cid:2)H O 3 3 3 2 2 Preparation as for [RhCl(dpq)([9]aneS )]Cl [28] [RhCl(phenylpy)([9]aneS )]Cl(cid:2)3H O 6 3 2 3 2 with 60.1 mg 2,3-diaminonaphthalene (0.38 mmol), dissolved in 10 ml CH Cl . Yield: 86.4 mg (40%), 500 ll 30% NaOMe (0.38 mmol) were added to 2 2 C H N Cl Rh S (cid:2)H O (565.82 g/mol) calcd. (%): 54.3 ll 2-phenylpyridine (0.38 mmol) and were 16 22 2 3 1 3 2 N 4.95 C 33.96 H 4.28 S 17.0, found (%): N 4.6 C stirred at 37(cid:3)C for 24 h. 100 mg of RhCl (cid:2)3H O 3 2 33.9H4.6S17.1;LSIMSm/z(%) = 511.0(34)[M- (0.38 mmol)dissolvedin30 mlMeOH/CH Cl (1/1) 2 2 Cl]?, 475 (78) [M-Cl-HCl]?, 439.1 (32) [M-Cl- was added and the reaction mixture was heated for 2HCl]?, 489.1 (100) [M-SCH CH ?2H]?, 1H-NMR 2 h in boiling MeOH/CH Cl After addition of 2 2 2 2. (DMSO-d , 400 MHz, 25(cid:3)C): d = 3.70–3.90 (m, 68.4 mg of [9]aneS (0.38 mmol), the resulting 6 3 12H, CH [9]aneS ), 7.50–7.60 (m, 2H), 7.90–8.05 suspension was refluxed for a further 5 h. Following 2 3 (m, 4H) ppm. filtration of the yellow precipitate ([RhCl ([9]ane 3 S )])andvolumereductionoftheremainingsolution 3 [RhCl(dab)([9]aneS )]Cl (cid:2)2H O 4 to 5 ml, addition of diethyl ether led to precipitation 3 2 2 oftheproduct,whichwaswashedwithH Oanddried 2 Preparationasfor[RhCl(bpy)([9]aneS )]Cl [28]with in vacuo. Yield: 25.2 mg (72%), C H NCl S Rh (cid:2) 3 2 17 20 2 3 1 41.6 mg2-phenylenediamine(dab)(0.38 mmol),dis- 3H O (562.4 g/mol) calcd. (%): N 2.49 C 36.31 H 2 solvedin10 mlMeOH/CH Cl .Yield:42.5 mg(21%), 4.66 S 17.1, found (%): N 2.45 C 36.51 H 4.61 S 2 2 C H Cl N S Rh(cid:2)2H O(533.79 g/mol)calcd.(%):N 16.45; LSIMS m/z (%) = 471.9 (100) [M-Cl]?, 12 20 3 2 3 2 123 658 Biometals(2011)24:645–661 437.1 (6) [M-Cl-HCl]?, 1H-NMR (D O ? NaCl, 5% CO for 72 h (MCF-7), or 48 h (HT-29 or HEK- 2 2 400 MHz, 25(cid:3)C): d = 3.4–3.75 (m, 12H, CH 293).InanalogytotheprocedureforMCF-7andHT- 2 [9]aneS ), 7.28–7.53 (m, 3H, H3/H7/H8), 7.77 (d, 29 cells, one plate was subsequently used for the 3 1H, H6), 7.95 (dd, 1H, H5), 8.12 (t, 1H, H4), 8.27 determination of the initial HEK-293 cell biomass. (d, 1H, H9), 8.74 (d, 1H, H2) ppm. Following removal of the medium, the HEK-293 cells were fixed by a 20–30 min incubation with 100 ll glutardialdehyde solution (1 ml glutardialde- Biological investigations hyde in 25 ml PBS, pH 7.3). The solution was removed,180 llPBSwereaddedandtheplatestored Cell cultures at 4(cid:3)C until further treatment. Stock solutions of the compounds 1–6 in DMF were freshly prepared and MCF-7 breast adenocarcinoma and HT-29 colon diluted with cell culture medium to the desired carcinoma cells were maintained in DMEM High concentrations (final DMF concentration: 0.1% v/v). Glucose (PAA) supplemented with 50 mg l-1 genta- Themediumintheplateswasreplacedwithmedium mycinand10%(v/v)fetalcalfserum(FCS)at37(cid:3)C/ containing the compounds at graded concentrations 5% CO and passaged once a week according to 2 (six replicates). After further incubation for 96 h standard procedures. For HEK-293 cells, Dulbecco’s (MCF-7), 72 h (HT-29) or 96 h (HEK-293), the modified eagle medium (Invitrogen) supplemented medium was removed and the plates treated with with 10% fetal bovine serum and 100 units/ml glutardialdehyde and PBS as described above. The penicillin and streptomycin was employed at 37(cid:3)C/ cell biomass was determined for HEK-293 and 5% CO . The cells were split and aliquots were 2 the other cells by crystal violet staining using the seeded in 35 mm culture dishes three times a week. following procedure. Firstly, the PBS was removed Burkitt-like lymphoma BJAB cells were maintained before addition of 100 ll of 0.02 M crystal violet at 37(cid:3)C in RPMI 1640 (GIBCO, Invitrogen) supple- solution and the plates were incubated for 30 min. mented with 10% heat-inactivated FCS, penicil- The crystal violet solution was subsequently lin (100,000 U l-1), streptomycin (0.1 g l-1) and removed, and the fixed cells were then washed twice L-glutamine (0.56 g l-1). The cells were subcultured with water and exposed to water for 15 min. The every 3–4 days by dilution of the cells to a concen- water was removed and the crystal violet extracted tration of 1 9 105 cells ml-1. Twenty-four hours fromtheadherentcellbiomassbyshakingfor3 hon before the assay setup, cells were cultured at a a softly rocking rotary shaker with 180 ll of 70% concentration of 3 9 105 cells ml-1 to ascertain ethanol. The absorption was then determined at standardized growth conditions. For the apoptosis 590 nm using a Fusion a multiwell-plate reader assays, the cells were then diluted to a concentration (Perkin–Elmer) for HEK-293 cells, or a Victor X4 of 1 9 105 cells ml-1 immediately before addition 2030 Multilabel Reader for MCF-7 and HT-29 cells. of the complexes. The mean absorption of the initial biomass plate was subtracted from the mean absorption of each exper- Antiproliferative activity measurements iment and control. IC values were determined as 50 those concentrations causing 50% inhibition of cell The antiproliferative effects of the compounds were proliferation. Results were calculated from 2 to 3 determinedbyanestablishedprocedurefortheMCF- independent experiments. 7 and HT-29 cells (Ott et al. 2005). An adaption of the procedure was employed for the HEK-293 cells Annexin-V propidium iodide binding assay (Bieda et al. 2010). Initial determination of a growth for Jurkat and K562 cells curve led to the HEK-293 specific parameters of cell concentration and incubation periods. Cells were Jurkat cells were purchased from German Collection suspended in cell culture medium (MCF-7: of Microorganisms and Cell Culture (DSMZ, Braun- 10000 cells ml-1; HT-29: 2,850 cells ml-1; HEK- schweig, No AA 282, LOT 7). The cells were 293: 2,000 cells ml-1) and 100 ll aliquots thereof maintained in RPMI 1640 (PAA) medium supple- were plated in 96 well-plates and incubated at 37(cid:3)C/ mentedwith10%fetalcalfserum(FCS,PAA),37(cid:3)C, 123 Biometals(2011)24:645–661 659 5% CO and maximum humidity. Jurkat/K562 cells (Becton–Dickinson) and CellQuest Pro (BD) analy- 2 were cultivated in standard conditions. sis software. Excitation and emission settings were After harvesting Jurkat cells/K562 cells and an 488 nm and 564–606 nm (FL2 filter), respectively. incubation period of either 24 or 48 h, the cells were The region of ‘low’- and ‘high’- ROS on histogram washed in cold phosphate-buffered saline (PBS). plot were determined according to the correspond- Negative controls were prepared by incubating cells ing control samples. in the absence of inducing agent. After resuspension at a concentration of 1 9 106 cells/ml in an annexin Cellular metabolism binding buffer (10 mM HEPES, 140 mM NaCl, and 2.5 mM CaCl , pH 7.4.), 100 ll of this suspension 2 Changes in cellular metabolism and morphology were prepared to provide a sufficient volume per were analyzed using a Bionas 2500 sensor chip assay. 5 ll of the annexin V conjugate (in 25 mM system(Bionas,Rostock,Germany).Thesensorchip HEPES, 140 mM NaCl, 1 mM EDTA, pH 7.4, plus (SC1000) enables continuous measurement of oxy- 0.1% bovine serum albumin (BSA); obtained from gen consumption using oxygen-sensitive electrodes Invitrogen) were added together with the dead-cell (Wolf et al. 1998), pH changes of the medium by indicator such as propidium iodide (SYTOX(cid:4) Green employing ion-sensitive field effect transistors dye) to each 100 ll of cell suspension. After an (Lehmann et al. 2000) and the impedance between incubationperiodof15 minatroomtemperature,the two interdigitated electrode structures (Ehret et al. cells were analyzed by flow cytometry using a 1998) to register the impedance under and across the FACS(cid:4)Calibur (Becton–Dickinson) and the Cell- cell layer on the chip surface. Before measurement, Quest Pro (BD) analysis software. cells were seeded on the sensor chip in DMEM (PAA, E15-883) with penicillin/streptomycin and PARP cleavage 10% (v/v) FCS (PAA) and incubated in a standard tissue culture incubator at 37(cid:3)C/5% CO and 95% Jurkatcellswereincubatedwithdifferentamountsof 2 humidityfor24 huntil90%confluencywasreached. selected compounds for 24 h. Cell lysing and Sensor chips with cells were then transferred to the analyzing were performed in a 10% SDS–PAGE. Bionas 2500 analyzer in which medium is continu- Proteins were transferred on Immobilon-P Transfer ously exchanged in 10 min cycles (3 min exchange membranes (Milipore) and the immune blotting was ofmediumand7 minwithoutflow)duringwhichthe performed with anti-PARP rabbits polyclonal anti- parameters were measured. The running medium bodies. For the signal detection a western blot used during the analysis was DMEM without car- lightning ECL-reagent and a LAS 3000 Imaging bonated buffer (PAN Cat. Nr. P03-0010) and only system (Fuji) were used. weakly buffered with 1 mM Hepes, reduced FCS (0.1%) and low glucose (1 g l-1). For drug activity Measurement of intracellular ROS testing, the following steps were followed: (a) 5 h towards the Jurkat and K562 cell lines equilibration with running medium with 3 and 7 min stop/flow incubation intervals, (b) 38 h drug incuba- Jurkat cells were cultivated in standard conditions tion with substances freshly dissolved in medium at during the treatment and the cells were treated with indicated concentrations also with the same stop/ indicated concentrations of the substances for 3 and flow,and(c)a9 hstepinwhichthecellswereagain 9 h. Cells were collected at given time points, fed with running medium without substances. At the centrifuged at 0.2 g (1,500 rpm) and resuspended in endofeachexperiment,cellswerekilledbyaddition FACS buffer (D-PBS, Gibco, ?1% BSA, PAA). of0.2%TritonX-100toobtainabasicsignalwithout The cell suspension was incubated with DHE living cells on the sensor surface with as a negative (dihydroethidium, SIGMA, 5 ll of 5 mM stock control. solution per 1 ml of cell suspension containing 106 cells) at room temperature in the dark for Acknowledgments Financial support for this work in 15 min., washed once more with FACS buffer and Bochum, Braunschweig and Heidelberg by the Deutsche immediately analysed using a FACS(cid:4)Calibur Forschungsgemeinschaft (DFG) within the research group 123 660 Biometals(2011)24:645–661 FOR630‘‘Biologicalfunctionoforganometalliccompounds’’is acid)(dppz)]n? (dppz=dipyrido[3,2-a:20,30-c]phenazine), gratefullyacknowledged.RuthBiedawishestothanktheRuhr n=1-3.JChemSocDaltonTrans,3664–3673 UniversityResearchSchoolfortheawardofascholarship. Harlos M, Ott I, Gust R, Alborzinia H, Wo¨lfl S, Kromm A, Sheldrick WS (2008) Synthesis, biological activity, and structure-activity relationships for potent cytotoxic rho- dium(III) polypyridyl complexes. J Med Chem 51: 3924–3933 References Hartinger CG, Zorbas-Seifried S, Jakupec MA, Kynast B, Zorbas H, Keppler BK (2006) From bench to bedside— Andree HA, Reutelingsperger CP, Hauptmann HC, Hemker preclinical and early clinical development of the anti- HC, Hermens WT, Willems GM (1990) Binding of vas- cancer agent imidazolium trans-[tetrachlorobis(1H-inda- cular anticoagulant-alpha (vac-alpha) to planar phospho- zole)ruthenium(III)] (KP1019 or FFC14A). J Inorg lipid-bilayers.JBiolChem265:4923–4928 Biochem100:891–904 Annen P, Schildberg S, Sheldrick WS (2000) HerebianD,SheldrickWS(2002)SynthesisandDNAbinding (g5-Pentamethylcyclopentadienyl)iridium(III) complexes properties of bioorganometallic (pentamethylcyclopentad ofpurinenucleobasesandnucleotides:acomparisonwith ienyl)iridium(III) complexes of the type [(C 5 Me 5 )Ir(- (g6-arene)ruthenium(II) and (g5-pentamethylcyclopenta- Aa)(dppz)]n? (dppz=dipyrido[3,2-a:20,30-c]-phenazine, dienyl)rhodium(III) species. Inorg Chim Acta 307:115– n=1-3), with S-coordinated amino acids (Aa) or pep- 124 tides.JChemSocDaltonTrans966–974 Bieda R, Ott I, Dobroschke M, Prokop A, Gust R, Sheldrick Jakupec MA, Galanski M, Arion VB, Hartinger CG, Keppler WS (2009a) Structure-activity relationships and DNA BK(2008)Antitumormetalcomplexes:morethantheme binding properties of apoptosis inducing cytotoxic rho- andvariations.DaltonTrans,183–194 dium(III) polypyridyl complexes containing the cyclic KapitzaS,JakupecMA,UhlM,KepplerBK,MarianB(2005) thioether[9]aneS .JInorgBiochem103:698–708 The heterocyclic ruthenium(III) complex KP1019 3 BiedaR,OttI,GustR,SheldrickWS(2009b)Cytotoxicrho- (FFC14A) causes DNA damage and oxidative stress in dium(III) polypyridyl complexes containing the tris(pyr- colorectaltumorcells.CancerLetts226:115–121 azolyl)methane coligand: synthesis, DNA binding Koopman G, Reutelingsperger CPM, Kuijten GAM, Keehnen propertiesandstructure-activityrelationships.EurJInorg RMJ,PalsST,vanOersMHJ(1994)Annexin-Vforflow Chem3821–3831 cytometric detectionofphosphatidylserine expression on Bieda R, Dobroschke M, Triller A, Ott I, Spehr M, Gust R, B-cellsundergoingapoptosis.Blood84:1415–1420 ProkopA,SheldrickWS(2010)Cell-selective,apoptosis- Lecoeur H (2002) Nuclear apoptosis detection by flow inducing rhodium(III) crown thiaether complexes. cytometry: influence of endogenous endonucleases. Exp ChemMedChem5:1123–1133 CellRes277:1–14 Brabec V, Kleinwa¨chter V, Butour J-L, Johnson NP (1990) Lehmann M,BaumannW, Brischwein M,EhretR, KrausM, Biophysical studies of the modification of DNA by anti- SchwindeA,BitzehoferM,FreudI,WolfB(2000)Non- tumor platinum coordination complexes. Biophys Chem invasivemeasurementofcellmembraneassociatedproton 25:129–141 gradientsbyion-sensitivefieldeffecttransistorarraysfor BrandtK,SheldrickWS(1996)Synthesis,structureandreac- microphysiologicalandbioelectronicalapplications.Bio- tivity of the thioether half-sandwich rhodium complex sensBioelectron15:117–124 [RhCl(MeCN) ([9]aneS )][CF SO ] .JChemSocDalton Levina A, Mitra A, Lay PA (2009) Recent developments in 2 3 3 32 Trans1237–1243 rutheniumanticancerdrugs.Metallomics1:458–470 Clarke M (2002) Ruthenium metallopharmaceuticals. Coord MadureiraJ,SantosTM,GoodfellowBJ,LucenaM,Pedrossade ChemRev232:69–93 JesusJ,Santana-MarquesMG,DrewMGB,FelixY(2000) Dobroschke M, Geldmacher Y, Ott I, Harlos M, Kater L, Structural characterisation of new Ru-II[9]aneS(3) poly- Wagner L, Gust R, Sheldrick WS (2009) Cytotoxic rho- pyridylcomplexes.JChemSocDaltonTrans4422–4431 dium(III) and iridium(III) polypyridyl complexes: struc- Marmur J (1961) Procedure for isolation of deoxyribonucleic ture-activity relationships, antileukemic activity, and acidfrommicro-organisms.JMolBiol3:208–218 apoptosisinduction.ChemMedChem4:177–187 Medvetz DA, Stakleff KD, Schreiber T, Custer PD, Hindi K, Ehret R, Baumann W, Brischwein M, Schwinde A, Wolf B PanznerMJ,BlancoD,TaschnerMJ,TessierCA,Youngs (1998)Onlinecontrolofcellularadhesionwithimpedance WJ (2007) Ovarian cancer activity of cyclic amines and measurements using interdigitated electrode structures. thiaethermetalcomplexes.JMedChem50:1703–1706 MedBiolEngComput36:365–370 MenonEL,PereaR,NavarroM,KuhnRJ,MorrisonH(2004) ErikssonM,NordenB(2001)Linearandcirculardichroismof Phototoxicityagainsttumorcellsandsindbisvirusbyan drug-nucleic acid complexes. Methods Enzymol 340: octahedral rhodiumbispyridyl complex and evidence for 60–98 the genome as a target in viral photoinactivation. Inorg Erkkila KE, Odom DT, Barton JK (1999) Recognition and Chem43:5373–5381 reaction of metallointercalators with DNA. Chem Rev Metcalfe C, Thomas JA (2003) Kinetically inert transition 99:2777–2795 metalcomplexesthatreversiblybindtoDNA.ChemSoc FrodlA,HerebianD,SheldrickWS(2002)Coligandtuningofthe Rev32:215–224 DNAbindingpropertiesofbioorganometallic(arene)ruthe- MorbidelliL,DonniniS,FilippiS,MessoriL,PiccioliF,Orioli nium(II) complexes of the type [(arene)Ru(amino P, Sava G,Ziche M(2003)Antiangiogenic properties of 123 Biometals(2011)24:645–661 661 selected ruthenium(III) complexes that are nitric oxide [L=Cl, n=1; L=(NH ) CS, n=2]. Eur J Inorg 22 scavengers.BrJCancer88:1481–1491 Chem3034–3046 Nicholson DW, Ali A, Thornberry NA, Vallancourt JP, Ding Scharwitz MA, Ott I, Geldmacher Y, Gust R, Sheldrick WS CK,GallantM(1995)Identificationandinhibitionofthe (2008) Cytotoxic half-sandwich rhodium(III) complexes: ice/ced-3 protease necessary for mammalian apoptosis. polypyridylligandinfluenceontheirDNAbindingprop- Nature376:37–43 erties and cellular uptake. J Organomet Chem 693: Ott I, Gust R (2007) Non platinum metal complexes as anti- 2299–2309 cancerdrugs.ArchPharmChemLifeSci340:117–126 SheldrickWS,Hagen-EckhardHS,HeebS(1993)Preparation OttI,SchmidtK,KircherB,SchuhmacherP,WiglendaT,Gust and structural characterization of tetranuclear (g6-ben- R (2005) Antitumor-active cobalt-alkyne complexes zene)ruthenium(II) complexes with bridging N-donor derivedfromacetylsalicyclicacid.Studiesofthemodeof atoms.InorgChimActa206:15–21 drugaction.JMedChem48:622–629 Shields TP, Barton JK (1995) Sequence-selective DNA rec- Pruchnik FP, Jakimowicz P, Ciunik Z, Zakrzewska-Czer- ognition and photocleavage—a comparison of enantio- winska J, Opolski A, Wietrzyk J, Wojdat E (2002)Rho- mersofRh(en)2Phi(3?).Biochemistry34:15037–15048 dium(III) complexes with polypyridyls and pyrazole and Simon H-U, Hai-Yehia A, Levi-Scha¨fer F (2000) Role of theirantitumoractivity.InorgChimActa334:59–66 reactive oxygen species (ROS) in apoptosis induction. Rademaker-LakhaiJM,VandenBongardD,PluimD,Beijnen Apoptosis5:418–425 JH,SchellensJHM(2004)AphaseIandpharmocological Sitlani A, Long EC, Pyle AM, Barton JK (1992) DNA study with imidazolium-trans-DMSO-imidazole-tetra- photocleavage by phenanthrenequinone diimine com- chlororuthenate—anovelrutheniumanticanceragent.Clin plexes of rhodium(III)—shape-selective recognition and CancerRes10:3717–3727 reaction.JAmChemSoc114:2303–2312 RiccardiC,NicolettiI(2006)Analysisofapoptosisbypropi- Van Engeland M, Nieland LJW, Ramaekers FCS, Schutte B, dium iodide staining and flow cytometry. Nat Protoc ReutelingspergerCPM(1998)AnnexinV-affinityassay:a 1:1458–1461 review on an apoptosis detection system based on phos- Rosenberg B, VanCamp L, Trosko JE, Mansour VH (1969) phatidylserineexposure.Cytometry31:1–9 Platinum compounds—a new class of antitumor agents. WangY,ZhangX,ZhangQ,YangZ(2010)Oxidativedamage Nature222:385 to DNA by 1, 10-phenanthroline/L-threonine copper(II) Sandstrom PA, Mannie MD, Buttke TM (1994) Inhibition of complexeswithchlorogenicacid.Biometals23:265–273 activation-induced death in T-cell hybridomas by thiol WolfB,BrischweinM,BaumannW,EhretR,KrausM(1998) antioxidants—oxidativestressasamediatorofapoptosis. Monitoring of cellular signaling and metabolism with JLeukocBiol55:221–226 modularservertechnique:thePhysioControlMicrosystem Scha¨ferS,OttI,GustR,SheldrickWS(2007)Influenceofthe (PCM).BiosensBioelectron13:501–509 polypyridyl (pp) ligand size on the DNA binding prop- ZeglisBM,PierreVC,BartonJK(2007)Metallo-intercalators erties, cytotoxicity and cellular uptake of organorutheni- andmetallo-insertors.ChemCommun4565–4579 um(II) complexes of the type [(C Me )Ru(L)(pp)]n? 6 6 123