← Back

Synthesis and DNA-binding properties of apoptosis-inducing cytotoxic half-sandwich rhodium(III) complexes with methyl-substituted polypyridyl ligands

JournalofOrganometallicChemistry696(2011)1023e1031 ContentslistsavailableatScienceDirect Journal of Organometallic Chemistry journal homepage: www.elsevier.com/locate/jorganchem Synthesis and DNA-binding properties of apoptosis-inducing cytotoxic half-sandwich rhodium(III) complexes with methyl-substituted polypyridyl ligands Yvonne Geldmachera, Riccardo Rubbianib, Pascal Wefelmeierc, Aram Prokopc, Ingo Ottb, William S. Sheldricka,* aFakultätfürChemieundBiochemie,Ruhr-UniversitätBochum,Universitätsstrasse150,D-44780Bochum,Germany bInstituteofPharmaceuticalChemistry,TechnischeUniversitätBraunschweig,Beethovenstr.55,D-38106Braunschweig,Germany cDepartmentofPediatricOncology,CologneChildren’sHospital,AmsterdamerStr.59,D-50735Cologne,Germany a r t i c l e i n f o a b s t r a c t Articlehistory: Half-sandwichorganorhodium(III)complexesofthetype[(h5-C5Me5)RhCl(pp)](CF3SO3)containingpoly- Received14July2010 pyridylligands(pp)representapromisingclassofcytostaticagents.Replacementofthepolypyridylligands Receivedinrevisedform ofcomplexes1(pp¼phen)and6(pp¼dppz)bymethyl-substitutedderivativesin2e5(pp¼4-Mephen, 1 A 5 cc O ep ct t o e b d e 1 r 8 2 O 01 c 0 tober2010 5-Mephen,4,7-Me2phen,5,6-Me2phen)and7(pp¼Me2dppz)leadstoasignificantimprovementintheir antiproliferativeactivitytowardshumanMCF-7andHT-29cancercells.Forinstance,theIC50valuetowards HT-29cellsdecreasesfrom4.3(cid:2)0.2mMfor6to0.98(cid:2)0.49mM forcomplex7.Incontrast,noactivity K R e h y o w di o u r m ds: (IC50 >100mM)wasobservedfortheHOOCandn-BuNHCOsubstituteddppzcomplexes8and9.UV/vis,CD andNMRspectraformixturesofcomplexes7e9withCTDNAwereinaccordancewithintercalationofthe Polypyridylligands Cytotoxicity substituteddppzligandsbetweenthebasepairsofthedoublehelixanddirectevidenceforthisbinding Apoptosis modewasalsoprovidedbya2DNOESYstudyforcomplex7withthehexanucleotided(50-CGTCGG-30).Each DNAbinding ofthemethyl-substitutedphencomplexes2e5issignificantlymoreactivetowardsimmortalizedHEK-293 cells(IC50values0.40(cid:2)0.02to0.94(cid:2)0.02mM)thantowardsthecancercells.Flowcytometricmeasure- mentsofDNAfragmentationinBJABcellsfollowinganincubationperiodof72hwith1,5and6indicatethat thecomplexesinducespecificapoptoticcelldeathinthenon-adherentlymphomacells. (cid:1)2010ElsevierB.V.Allrightsreserved. 1. Introduction reported for complexes of the type [(h5-C5Me5)MCl(pp)](CF3SO3) (M¼Ir,pp¼dppz,dppn[3,6];M¼Rh,pp¼phen,dpq,dppz,dppn Investigationsofthecytotoxicpropertiesofareneruthenium(II) [7])containingthe polypyridylligands1,10-phenanthroline(phen) complexes with amino acidato [1], diamine [2], diimine [3] and dipyrido[3,2-f:20,30-h]quinoxaline (dpq), dipyrido[3,2-a:20,30-c] phosphaneligands[4]haveestablishedpromisinganticanceractivity phenazine (dppz) and benzo[i]dipyrido[3,2-a:20,30-c]phenazine for this class of half-sandwich compounds. In striking contrast, it (dppn)(Fig.S6).Acleardependenceonthesizeofthepolypyridyl is only very recently that similar pentamethylcyclopentadienyl ligand is apparent for the cytotoxicity of these rhodium(III) complexesoftheneighbouringGroup9transitionmetals,rhodium compoundstowardsthehumanMCF-7(breastcarcinoma)cellline. (III)andiridium(III),haveattractedanyinterestinthisrespect.Afirst TheirIC50valuesdecreaseintheligandorderphen,dpq>dppz> articlebyDysonetal.in2006[5]reportedaninvitroevaluationofthe dppnfrom4.7/5.1over1.5to0.8mM[7].Markedinvitrocytotoxicity RAPTA [4] analogues [(h5-C5Me5)RhCl2(pta)] and [(h5-C5Me5) has also recently been reported for analogous (h5-C5Me5)M(III) RhCl(pta)2] þ (pta¼1,3,5-triaza-7-phosphatricyclo[3.3.1.1]decane) (M¼Rh,Ir)complexescontainingthechelatingligands2-(pyridine- towardsthehumancelllinesHT-29(coloncarcinoma),A549(lung 2-yl)thiazole (towards the A2780 and A2780cisR cell lines) [8] carcinoma)andT47D(breastcarcinoma),butestablishedonlyvery and 1,2-naphthoquinone-1-oximate (towards the HeLa and HL60 limited cytotoxicity, as indicated by the determined IC50 values of celllines)[9]. greater than 380mM. A much higher activity was subsequently CelldeathinducingpropertiesforthehumancancercelllinesMCF- 7andHT-29(coloncarcinoma)havealsobeenestablishedforrhodium (III)complexesofthetype[(coligand)RhCl(pp)](CF3SO3)containingthe * C E- o m rr a e i s l p a o d n d d re in ss g : a w u i t l h li o a r m .T .s e h l. e : ld þ r 4 ic 9 k@ 23 ru 4 b 3 .d 2 e 24 ( 1 W 9 . 2 S ; . f S a h x e : ld þ r 4 ic 9 k) 2 . 343214420. facial coligands 1,4,7-trithiacyclononane ([9]aneS3) [10] and tris 0022-328X/$eseefrontmatter(cid:1)2010ElsevierB.V.Allrightsreserved. doi:10.1016/j.jorganchem.2010.10.034 1024 Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 Scheme 1. Cations and the employed numbering scheme of the compounds [(h5-C5Me5)RhCl(pp)](CF3SO3)1e5withphenandsubstitutedphenanthrolineligands. (pyrazolyl)methane(tpm)[11].Severalinterestingstructureeactivity relationships are apparent on comparison of the IC50 values for rhodium(III) complexes containing different tridentate coligands. In particular,theorderoftheIC50values(hereforMCF-7cells)changesin Fig.1. Molecularstructureofthecationofthe5-methylphenanthrolinecomplex3. dependenceonthecoligandfromtpm>[9]aneS3 >[h5-C5Me5] (cid:3) for pp¼phen (51.7>36.3>4.7mM) over [9]aneS3 >[h5-C5Me5] (cid:3) , tpm the cell death induced by these half-sandwich complexes. In forpp¼dpq(19.1>5.1,4.0mM)to[9]aneS3 >[h5-C5Me5] (cid:3)>tpmfor contrast to ([9]aneS3)Rh(III) half-sandwich complexes and tri- pp¼dppz (4.7>1.5>0.43mM). Based on these results, it can be chlorido-rhodium(III) complexes of the type mer-[RhCl3(DMSO) concludedthattheextentoftheinfluenceofpolypyridylligandsizeon (pp)], which induce specific apoptotic cell death via the intrinsic compound cytotoxicity increases in the order of the coligands [h5- mitochondrialpathway[10,16],thepronouncedcytotoxicityof(h5- a C t 5 t M em e5 p ] t (cid:3) s < t [ o 9]a i n m e p S r 3 o < ve tpm the [11 cy ]. to W to e x , ic t i h ty ere o fo f re [ , (c c o o li n g c a l n u d d ) e R d hC th l( a p t p) a ] n nþ y C m 5 a M te e5 a ) s R a h( b II i I d ) en c t o a m te p l l i e g x a e n s d c h o a n s ta b i e n e i n ng att 1 r , i 2 b - u n t a e p d h t t o ho n q e u cr in o o si n s e [ - 9 1 ] - . o W xi e - compounds(n¼1,2)containingderivativesofthesmallerpolypyridyl nowreportstudiesofpossiblenecrosisand/orapoptosisinduction ligand 1,10-phenanthroline should concentrate on employing pen- in lymphoma BJAB cells following their incubation with the tamethylcyclopentadienylasthefacialcoligand. complexes1(pp¼phen),5(pp¼5,6-Me2phen),6(pp¼dppz)and Onepromisingstrategyforenhancingthebiologicalactivityof 8(pp¼HOOC-dppz). phenanthroline complexes is to replace phen by its methyl- substituted derivatives, as first demonstrated by McFadyen et al. [12] for [Pt(en)(3,4,7,8-Me4phen)]2þ (IC50 ¼0.7mM) against the 2. Resultsanddiscussion murine leukemia L1210 cell line in comparison to [Pt(en)(phen)] 2.1. Synthesisof1e9 (IC50 ¼2mM). This approach was subsequentlyextended to mon- omethyl- (4-Mephen, 5-Mephen) and dimethyl-substituted phe- The half-sandwich complexes [(h5-C5Me5)RhCl(pp)](CF3SO3) n [1 a 3 n ,1 th 4] r , o w lin h e o s d ( i 4 s , c 7 o - v M er e e 2 d ph th en at ,5 th ,6 e - 5 M ,6 e - 2 M ph e e 2 n p ) he b n y c A o ld m r p ic l h ex -W ex r h ig i h bi t ts et th a e l. 1 R e hC 9 l( w ac e e re to p n r e e ) p 2] a ( r C e F d 3S b O y 3) re w flu it x h in t g he th a e p s p o r l o v p e r n i t at c e om po p l l y e p x y [ r ( i h dy 5- l C l 5 ig M a e n 5 d ) h th ig e h se es P t t( a I c I) ti c v o it m y. p V le i x s e co s m ar e e tr D y N a A n i d nt C e D rc / a L l D at s o t r u s d a i n es d i t n h d a i t c t a h te e d pr t e h s a e t n a c l e l (pp) in CH3OH/CH2Cl2 (1:1) for 5h. [(h5-C5Me5)RhCl(acetone)2] þ d of iffe m re e n th c y e l s i s n ub t s h t e itu ge e o n m ts et a r t y v o a f r t y h in ei g r D p N os A iti b o i n n s din c g au t s h e a s tm sig a n y ifi af c f a e n ct t R w a h a C s s o l} l o 2 u b ( t m i t o - a C n in l) e 2 o d ] f [ i 1 n t 7 h ] e si i t n d u in a b c u y e c t l o e a a n d r e dit a s i t n o a n d rti s o n u f g b 2 se c e o q q m u u e p i n v o t u o c n f e d n A t g r [{ i ( f ( C u h F g 5 3 a - S C t O i 5 o 3 M n ) e t o 5 o ) f theircytotoxicity[13]. precipitated AgCl after stirring in the dark for 2h. Reaction of We have now investigated the biological activity and DNA- dipyrido[3,2-a:20,30-c]phenazine-7-carbonic acid (HOOC-dppz) bindingpropertiesofthecomplexes2e5(Scheme1)withmethyl- [18]withn-butylaminefor2hat80(cid:4)Cinapyridinesolutioninthe substituted phenanthroline ligands in comparison to[(h5-C5Me5) presence of the coupling agent 1,10-carbonyldiimidazole led to RhCl(phen)](CF3SO3)1[7],andthecomplexes7e9(Scheme2)with formation of the novel ligand (7-n-butylamido)dipyrido[3,2- substituteddppzligandsincomparisonto[(h5-C5Me5)RhCl(dppz)] a:20,30-c]phenazine(n-BuNHCO-dppz),whichwascharacterizedby (CF3SO3)6[15].Asecondpointofinterestwasthemechanismof elemental analysis, LSIMS (liquid secondary ion mass spectrom- etry)and1HNMRspectroscopy,aswerecomplexes1e9. The structures of compounds 3 and 5 were determined by X-raystructuralanalysisandtheirorganorhodium(III)cations are depicted in Figs. 1 and 2. Crystal and refinement data are summarizedinTable1.Bondlengthsandanglesinthecationsof 3 and 5 are similar to those previously reported for [(h5-C5Me5) RhCl(pp)]X(pp¼bpy,phen;X (cid:3)¼ClO4 (cid:3) ,CF3SO3 (cid:3) )[7,19,20]and[(h5- C5Me5)RhCl(pp)](CF3SO3)(pp¼dpq,dppz,dppn)[7].Forinstance, the 5,6-Me2phen complex 5 exhibits RheN distances of 2.102(2) and 2.121(2) (cid:2) A, RheC distances in the range 2.151(2)e2.180(2) (cid:2) A andanRheClbondlengthof2.398(1) (cid:2) A.Theanalogousdistances in [(h5-C5Me5)RhCl(phen)](CF3SO3) [7] are 2.100(2) and 2.121(2), 2.141(3)e2.171(3) and 2.406(1) (cid:2) A. Pronounced envelope confor- mationsareadoptedbythefive-memberedNRhNCCchelaterings in both 3 and 5, as can be gauged by the respective interplanar angles of 8.3(2)(cid:4) and 10.3(2)(cid:4) between their N1eRh1eN10 and N1eCeCeN10 planes. Contrasting angles of 14.7(4) and 1.7(5)(cid:4) Scheme 2. Cations and the employed numbering scheme of the compounds [(h5-C5Me5)RhCl(pp)](CF3SO3)6e9withdppzandsubstituteddppzligands. were observed for the analogous compounds [(h5-C5Me5)RhCl Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 1025 Fig.2. Molecularstructureofthecationofthe5,6-dimethylphenanthrolinecomplex5. Fig.3. CDspectraofCTDNAandmixturesofcomplexes3and5withCTDNA(r¼0.2) ina10mMphosphatebuffer(pH7.2)aftera2hincubationperiod.Molarellipticities (pp)](CF3SO3) with the larger polypyridyl ligands dpq and dppz [q]aregivenintheunitsdegcm2dmol(cid:3)1(cid:5)10(cid:3)3. [7], which exhibit interplanar angles of 76.6(2) and 58.5(3)(cid:4) 1e5at1:5complex/DNAmolarratios([DNA]¼M(basepairs))leads between their cyclopentadienyl and polypyridyl ring systems. Somewhat lower interplanar angles of 50.1(2) and 49.4(2)(cid:4) are torelativelyminorchangesinthel maxvaluesandmolarellipticities [q]oftheseCDbands.Fig.3depictstheCDspectraformixturesofthe foundinthecationsof3and5,respectively. 5-Mephen and 5,6-Me2phen complexes 3 and 5 with CT DNA at r¼0.2(r¼[complex]/[DNA]).Itisapparentthattheinteractionwith 2.2. DNA-bindingstudies thecomplexesleadstoabathochromicshiftforl maxofthepositive band to about 280nm in both cases. Whereas complex 3 causes Stable intercalative binding into DNA has previously been a small increase in the molar ellipticity [q] of this band, a 10% confirmed by induced CD bands and large viscosity increases for decrease is observed in the presence of 5. Similar changes were half-sandwich rhodium(III) polypyridyl complexes with pp¼dpq recordedforcomplex/DNAmixturescontainingthe4-Mephenand anddppz(6)[7].Asboththesecomplexesarehighlycytotoxicitis 4,7-Me2phencomplexes2and4(Fig.S1).Bathochromicshiftsand plausible that their activity may be related to DNA distortion or smalldecreasesinmolarellipticityhaveoftenbeenregisteredforCT damage caused by this binding interaction. It was, therefore, of DNAspectrafollowingtheformationofcovalentadductsthrough interest to study the effects of both electron-donating methyl metalenucleobasebinding[3].Incontrast,thesignificantincreases groups(complexes2e5,7)andelectron-withdrawingCOOHandn- inmolarellipticityandpronouncedblueshiftsforthepositiveCD BuNHCOgroups(complexes8and9)ontheDNAbindingofsuch bands of complex/DNA mixtures (r¼0.2) containing 7e9 (Fig. 4) organometallicrhodium(III)complexes. maybeattributedtochangesinnucleobasestackingduetopoly- pyridylligandintercalation[3,6,7].Additionalstrongevidenceforan 2.2.1. Electronicspectra intercalativebindingmodeisprovidedbythepresenceofinduced AnegativeCDbandat246nmduetothehelicalBconformation negativeCDbandsataboutl¼296nm.Thelowermolarellipticity and a positive band at 275nm caused by nucleobase stacking of(cid:3)2.7(cid:5)10 (cid:3)3(l¼297nm)forthemixtureofcomplex8withCT representthecharacteristicfeaturesinthecirculardichroism(CD) DNA suggests, in comparison to the [q] values of (cid:3)4.6(cid:5)10 (cid:3)3 spectrumofcalfthymusDNA(CTDNA)[21].Additionofcomplexes (l¼301nm)and(cid:3)3.9(cid:5)10 (cid:3)3degcm2dmol (cid:3)1(l¼296nm)forthe other complexes 7 and 9 (Table 2), that the interaction may be Table1 Crystalandrefinementdatafor3and5. weakerinthecaseoftheHOOC-dppzcompound8.A[q]valueof (cid:3)4.5(cid:5)10 (cid:3)3degcm2dmol (cid:3)1 (l¼296nm), very similar to that Compound 3 5 observedfortheMe2dppzcomplex7,waspreviouslyreportedfor Formula C24H25ClF3N2O3RhS C25H27ClF3N2O3RhS theunsubstituteddppzcomplex6[7]. M[gmol(cid:3)1] 616.9 630.9 ThermaldenaturationmeasurementsforCTDNAinthepresence T[K] 115 115 Radiation MoKa MoKa oftransitionmetalcomplexesprovideasimplemeansofgauging Crystalsystem Triclinic Triclinic Spacegroup P1 P1 (cid:2) a[A] 8.0012(4) 7.9010(3) (cid:2) b[A] 11.8641(6) 12.1663(8) (cid:2) c[A] 13.1047(6) 13.2542(9) a[(cid:4)] 87.742(4) 90.180(6) b[(cid:4)] 79.559(4) 100.962(4) g[(cid:4)] 81.405(4) 98.884(4) V( (cid:2) A3) 1209.56(10 1235.13(13) Z 2 2 D[gcm(cid:3)3] 1.694 1.696 F(000) 624 640 m[mm(cid:3)1] 0.956 0.938 2q max[(cid:4)] 50.0 57.5 Collectedrefl. 6974 11,906 Independentrefl. 4263 5517 Refinedparameters 328 332 R1[I>2s(I)] 0.031 0.027 wR2(alldata)a 0.057 0.059 S(goodness-of-fit) 0.800 0.920 Max./min.Dr[e (cid:2) A(cid:3)3] 0.54/(cid:3)0.44 0.56/(cid:3)0.51 Fig.4. CDspectraofCTDNAandmixturesofcomplexes7e9withCTDNA(r¼0.2)in a10mMphosphatebuffer(pH7.2)aftera2hincubationperiod.Molarellipticities[q] a wR2 ¼[Sw(Fo 2(cid:3)Fc 2)2/Sw(Fo 2)2]1/2. aregivenintheunitsdegcm2dmol(cid:3)1(cid:5)10(cid:3)3. 1026 Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 Table2 DA/Avaluesof(cid:3)47,(cid:3)44and(cid:3)44%fortheabsorptionbandatabout Characteristicparametersfor theinteractionofcomplexes6e9withCTDNA in 370nmbelongingtocomplexes6,8and9arethereforetypicalfor a10mMphosphatebufferatpH7.2. thisbindingmodeandareaccompaniedbyredshiftsof5,4 and Complex DTm a((cid:4)C) [q]b DA/Ac(%) 4nm, respectively. In contrast, a much lower decrease in absor- 6 þ12 (cid:3)4.5(cid:5)10(cid:3)3 (cid:3)47 banceof(cid:3)23%withanassociatedredshiftof5nmisobservedfor 7 8 þ þ 7 4 (cid:3) (cid:3) 4 2 . . 6 7 (cid:5) (cid:5) 1 1 0 0 (cid:3) (cid:3) 3 3 (cid:3) (cid:3) 2 4 3 4 t in he th a e na o l u o t g e o r u b s e b n a z n e d ne of ri t n h g e l M ea e d 2 s d t p o p i z n c c o re m a p se le d x e 7 le .M ctr e o th n y d l e s n u s b i s t t y it i u n t t io h n e 9 þ8 (cid:3)3.9(cid:5)10(cid:3)3 (cid:3)44 dppzringsystemandalsocausesasignificantdisruptionofsolvent a Meltingtemperatureshift. hydrogen bonding to the quinoxaline nitrogen atoms [23]. It is, b Molarellipticity(degcm2dmol(cid:3)1)ofthenegativeICDataboutl¼296nm. therefore,possiblethatsolventeffectsorchangesinthepolypyridyl c Absorptionshiftofthecomplexabsorptionmaximumataboutl¼370nm. ligandorientationrelativetothebasestackmayberesponsiblefor anystabilizationofthedoublehelixduetoabindinginteraction. the much lower decrease in absorption at the 370 and 390nm Whereas the intercalation of polypyridyl ligands between the maximaoftheMe2dppzcomplex7followingitsintercalationinto duplexnucleobasepairsgenerallyleadstolargeincreasesintheDNA the double helix. An analogous trend has been reported for the meltingtemperature,covalentmetalenucleobasebindingtypically complexcations [Ru(phen)2 (Mendpq)]2þ (n¼0,2). Whereas DNA causesonlyminorchangesinTmandcaneveninvokeadestabiliza- interaction causes a hypochromic shift DA/A of (cid:3)21% for n¼0, tionoftheBconformation[7].TheDTmvalueslistedforcomplexes amuchsmallerDA/Ashiftofonly(cid:3)7%wasobservedforn¼2[24]. 6e9 in Table 2 were recorded under equilibrium conditions in a 10mM phosphate buffer at pH 7.2 for 1:5 complex/CT DNA 2.2.2. Viscositymeasurements mixturesandareindicativeofDNAintercalationforthecationsof ThemostconvincingevidenceforDNAintercalationisprovided the substituted dppz complexes 7 and 9. It is interestingto note, by viscosity titrations [25,26]. Insertion of extended aromatic however,thattheDTmvalueofþ12(cid:4)Cfortheunsubstituteddppz ligands such as dppz between adjacent nucleobase pairs causes complex6issignificantlyhigherthanthevaluesdeterminedforthe lengtheningandstiffeningofthedoublehelixandsuchchangesare substituteddppzcomplexes.ThissuggestsweakerDNAbindingfor reflected in an increase in DNA viscosity. The reduced viscosity 7e9duepossiblytostericcontactstothesubstitutedgroupscausing function(h/h 0)1/3isequalto(L/L0),wherehandLarethereduced alessfavourablepolypyridylligandorientationrelativetothebase viscosityandthelengthoftheDNAatagivencomplex/DNAratio pairs. The lower DTmvalue of þ4(cid:4)C for the HOOC-dppz complex r,andh 0andL0arethecorrespondingvaluesforDNAalone[25].At 8 correlates well with the lower [q] value of (cid:3)2.6(cid:5)10 (cid:3)3 low complex/DNA ratios (0(cid:6)r(cid:6)0.15), ideal monointercalation degcm2dmol (cid:3)1 for its negative ICD at l¼297nm. Although shouldleadtoavaluecloseto1.0forthefunction(L(cid:3)L0)/(L0 $r),on theincreasesintheDNAmeltingtemperatureforthepresumably thebasisoftheclassicalmodelofintercalation,wherethehelixis non-intercalatingcompounds2e5aremuchlowerthanfor6,7and lengthenedbyabout3.4 (cid:2) Aperintercalatedmoiety.Infact,amuch 9,theirrespectivevaluesofþ4,þ3,þ5andþ4(cid:4)Careatthesame widerrangeofhelixextensionsfrom2.0 toabout3.7 (cid:2) A hasbeen timesignificantlylargerthanthatof(cid:3)1(cid:4)Cpreviouslyrecordedfor established for aromatic molecules by electric dichroism the phen complex 1 [7]. This suggests that hydrophobic contacts measurements [27]. On plotting experimental (h/h 0)1/3 values involvingthemethylsubstituentsmaypossiblyhaveastabilizing against the complex/DNA ratio r for r(cid:6)0.15, slope values for influenceontheCTDNAdoublehelix. (L(cid:3)L0)/(L0 $r) of 0.71,1.02 and 0.93 were obtained for complexes As depicted in Fig. 5 for complex 9, the (h5-C5Me5)Rh(III) 7, 8 and 9, respectively. All three values are indicative of strong complexesofsubstituteddppzligandsexhibitcharacteristicbands intercalativebinding. atabout370and390nm,whichareduetopep*transitioncentred inthephenazinepartof theligand[22].Followinganincubation 2.3. NOESYstudies periodof2min,UV/visspectraofbufferedsolutionsofcomplexes 6e9(10mMphosphatebuffer,pH7.2)withCTDNAat25(cid:4)Cand ThebindingmodeoftheMe2dppzcomplex7toaB-typedouble r¼0.2 exhibit no further significant changes over the next 24h, helixwasalsoinvestigatedby1HNMRspectroscopy.Aminimumof thereby indicating that achievement of equilibrium conditions six base pairs is necessary to enable the adoption of a normal mustberelativelyrapid.Pronounceddecreasesinabsorbancefor B-typedoublehelixwithamajorandminorgroove[28].Thishas themaximaatabout370and390nmandbathochromicshiftsof led to the palindromic hexanucleotide d(50-GTCGAC-30)2 being these maxima are generally indicative of intercalation of dppz often employed to study the interaction of metal-containing ligandsbetweenthenucleobasepairsofthedoublehelix[6,7].The monointercalators with a model system for DNA [15,29,30]. Both the possibility of competition between a number of preferred intercalative binding sites and extensive peak overlapping can prevent a detailed 2D NMR spectroscopic analysis of interactions withlongeroligonucleotides [30].Theemploymentofnon-palin- dromicoligonucleotidesavoidstheassignmentambiguitiescaused bysymmetry-relatedbindingssites[31]andwe,therefore,chosed (50-CGTCGG-30)forourNOESYstudies.Thishexanucleotideoffers pyrimidineepyrimidine, pyrimidineepurine and purineepurine sequences for intercalative binding and was selected to contain a GTC sequence on the basis of the previous observation of respectively T2/C3 and G1/T2 intercalation for D-[Ru(phen)2 (dpq)]2þ [30]and[(h5-C5Me5)Ir(dppz)(H2-Met-OMe)]3þ [15]with palindromicd(50-GTCGAC-30)2. 1H NMR analysis of changes in the exchange kinetics of the hydrogen-bonded nucleobase imino protons on metal complex bindingtoDNAoftenprovidesusefulinsightsintothenatureofthe Fig.5. UV/visspectrumofthen-BuNHCO-dppzcomplex9anditsinteractionmixture withCTDNAatr¼0.2ina10mMphosphatebuffer(pH7.2)after2min. interaction with the double helix [32,33]. Upfield shifts and Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 1027 Fig.6. Changesintheiminoprotonregion(12.2e14.0ppm)ofthe1HNMRspectrum at288Kofd(50-CGTCGG-30)uponadditionofthecomplex[(h5-C5Me5)RhCl(Me2dppz)] Cl(7)atr¼0.2,0.4,0.6and1.0. Fig.7. Two-dimensionalNOESYspectrumofthearomaticMe2dppzproton/nucleobase H6,H8andsugarH20,H200þT3-CH3regionofthe1.1mixtureof7andd(50-CGTCGG- 30)atpH7.2and288K.ThenucleotidesarenumberedfromC1toG6inthefirststrand pronouncedpeakbroadeningarecharacteristicforcompoundsthat andC7toG12inthesecondstrand. intercalateintotheduplex[34]andbothphenomenaareobserved fortheG2,T3,G5andG9iminoprotonsformixturesofcomplex Inthesecases,NOEcrosspeakswereinaccordancewithdppzinsertion 7andd(50-CGTCGG-30)atr¼0.2,0.4and0.6(Fig.6).Anassignment betweenA5/T8andC6/G7. of theGiminoprotonswasnotpossibleowingtotheabsenceof suitable NOE cross peaks. The signals for the T3 and two of the 2.4. Cytotoxicitymeasurements G imino protons fully disappeared at r¼1.0 leaving only a very broad resonance for the remaining G imino proton. A tentative Invitrocytotoxicitystudiesofcompounds1e9wereperformed assignmentofthisresonancetoG2ispossibleonthebasisofthe withtheadherentcancercelllinesMCF-7andHT-29andwiththe lowerdegreeofbroadeningobservedforG2eH6incomparisonto adherent immortalized cell line HEK-293 (human embryonal G5eH6andG9eH6.Thewideningoftheiminoprotonsignalsisin kidney).TheresultingIC50valuesarelistedinTable3.Stocksolu- accordance with rapid solvent exchange due to weakening or tionsofthecompoundsinDMFwerefreshlypreparedanddiluted interruption of the WatsoneCrick hydrogen bonds. Both the with cell culture medium prior to incubation for 72h (HT-29) or broader signals and the upfield shifts are clearly indicative of 96h (MCF-7 and HEK-293) with the cells. Reproducible results intercalativebinding. could notbe obtained for HT-29 cells in the presence of the 5,6- Both dppz/nucleotide and Cp*eCH3 /nucleotide NOEs could Me2phen complex, presumably owing to the relatively low solu- be detected in the two-dimensional NOESY spectrum of the bilityofthishalf-sandwichcompoundinthecellculturemedium. complex7/d(50-CGTCGG-30)mixtureat r¼1.0(Fig.7).The nucle- Anincreaseintheantiproliferativepropertiesofthe(h5-C5Me5)Rh obase H6/H8 resonances were assigned on the basis of intra- (III)complexesonreplacementofthephenordppzligandsof1and molecular nH20/and nH200/(nþ1) H6/H8 cross peaks. An 6bymethyl-substitutedderivativesisapparentforcomplexes2e5 upfieldshiftto8.95/8.99ppmwasobservedfortheMe2dppz-H2/9 and7.IncontrasttothefindingsofAldrich-Wrightetal.forPt(II) protonsatr¼1.0inPBSincomparisontothedoubletat9.29ppm complexesofthetype[Pt(en)(pp)]Cl2[13],thesechangesarenow forthecomplexonitsowninCD2Cl2.Nosignalscouldbeidentified morepronouncedforthe4-Mephenand4,7-Me2phencompounds. for the Me2dppz-H4/H7 protons at r¼1.0, presumably due to Forinstance,the4-Mephenand4,7-Me2phencomplexes2 and4 pronouncedbroadeningcausedbyintercalativebinding.Addition aresome2.6and3.5timesmoreactivetowardsHT-29cellsthan of the Me2dppz complex to the hexanucleotide also induced thephencomplex1,whereasonlyatwofoldincreasewasobserved significantwideningofthearomaticH6/H8resonancesoftheC4, forthe5-Mephencomplex3.Inthe(en)Pt(II)series,theIC50values G5, G6, C7 and C8 nucleobases, which suggests that the metal towards murine leukemia L1210 cells improve from 9.7(cid:2)0.4mM complexmaypreferentiallyintercalatebetweenbasepairsinthis over 2.8(cid:2)0.8 to 1.5(cid:2)0.3mM for pp¼phen, 5-Mephen, 5,6- partoftheoligonucleotide. TheintermolecularNOEcrosspeakslinkingG6eH8(7.60ppm)and Table3 C7eH6(7.53ppm)withtheCp*eCH3protonsofcomplex7(1.65ppm) IC50values(mM)aforthecomplexes[(h5-C5Me5)RhCl(pp)](CF3SO3)1e9towardsthe andtheNOEbetweenthesugarC7eH20proton(1.98ppm)andtheH2/ celllinesMCF-7,HT-29andHEK-293. 9protons(8.95,8.99ppm)ofitsdppzligand(Fig.7)areinaccordance Compound pp MCF-7 HT-29 HEK-293 withthisconclusion.Thesecontactsprovideevidencefordppzinter- 1 Phen 4.7(0.01) 8.0(1.6) n.d. calationbetweenthebasepairsG5/C8andG6/C7.TheweakerNOE 2 4-Mephen 1.95(1.30) 3.1(1.1) 0.94(0.02) betweentheH2/H9protonsandC11eH20(1.74ppm)andthecontact 3 5-Mephen 3.79(0.25) 3.96(0.15) 3.18(0.06) betweenC1eH6(7.43ppm)andtheCp*eCH3protonsindicatethat 4 4,7-Me2phen 2.50(0.28) 2.27(1.80) 0.40(0.02) aminorintercalationmodebetweenC1/G12andG2/C11mayalsobe 5 5,6-Me2phen 3.97(0.04) n.d. 0.85(0.14) 6 dppz 1.5(0.4) 4.3(0.2) 2.80(0.10) involved.Suchasequence-selectiveintercalationbetweenthefirst/last 7 Me2dppz 0.93(0.09) 0.98(0.49) 2.48(0.69) two base pairs of the octanucleotide would be analogous to that 8 HOOC-dppz >100 >100 >100 previously reported for [(h5-C5Me5)Ir(dppz)(H2-Met-OMe)]3þ and 9 n-BuNHCO-dppz >100 >100 >100 [(h6-C6Me6)Ru(Ac-Met-OH)(dppz)]2þ withd(50-GTCGAC-30)2[15,35]. a Thevaluesinbracketsrepresentthestandarderrors;n.d.¼notdetermined. 1028 Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 Fig.8. InhibitionofcellproliferationinBJABcellsaftertreatmentwithcomplexes5(a) and6(b)for24hasmeasuredbyaCASYcellcounter.N¼numberofcellsinunitsof Fig. 9. DNA fragmentation of BJAB cells after incubation for 72h with different 105cells/ml(cid:2)ESD(n¼3);I,thepercentageinhibitionofcellproliferationisdefinedas concentrationsofcomplexes5(a)and6(b).Dataaregivenin%hypoploidy(subG1) I¼[{(Nc)t (cid:3)(Ns)t}/{(Nc)t (cid:3)(Nc)0}](cid:5)100%, where (Nc)0 and (Nc)t are the numbers of whichreflectsthenumberofapoptoticcells. c o e f l c ls el i l n s t i h n e a co sa n m tr p o l l e be fo fo ll r o e w a i n n d g a tr ft e e a r tm th e e n 2 t 4 w h it i h nc a u c b o a m tio p n le p x e a ri t o a d g a i n v d en (N c s o ) n ti c s e t n h t e ra n ti u o m n b fo e r r complex6towardsHEK-293cellsissimilartothevaluesof4.3(cid:2)0.2 24h. and1.5(cid:2)0.4recordedforHT-29andMCF-7cells. M ina e c 2 t p i h ve en ( , IC b 5 u 0 t > bo 5 t 0 h m t M he ) 4 [ - 1 M 3] e . p I h t e s n ho a u n l d d, 4 h ,7 o - w M e e v 2 e p r h , e b n e c n o o m te p d lex th e a s t ar in e 2.5. Apoptosisinduction contrastto2e5theplatinum(II)complexesareDNAintercalators andthatCDmeasurementswereconsistentwithdifferingconfor- Compounds 1, 5 (pp¼phen, 5,6-Me2phen) and 6 (pp¼dppz) were selected as biologically active compounds to ascertain m [13 at ]. io T n h a e l c c y h to an to g x e i s cit w y h r e e n su t l h ts e f s o e r c 2 o e m 5 p in o d u i n c d a s te i t n h t a e t ra b c o t th w t i h th ep C o T si D ti N o A n whetherthecelldeathinvokedby(h5-C5Me5)Rh(III)complexesis duetoapoptosisand/ornecrosis.Thecomplex8,whichisinactive (s)andthenumberofmethylsubstituentshavelessinfluenceon towards MCF-7 and HT-29 cells, was studied for comparison theactivityoftheorganometallicrhodium(III)complexes.Eachof purposes.Necroticcelldeathischaracterizedbyanearlyreleaseof the methyl-substituted phenanthroline complexes 2e5 is signifi- lactatedehydrogenase(LDH),whereasapoptoticcellsretaintheir cantly more active towards HEK-293 cells in comparison to the membrane integrity and do not exhibit an early release of large cancercelllinesMCF-7andHT-29. intracellularproteinsincludingLDH[37].NosignificantLDHrelease Respectively 4.4- and 1.6-fold improvements in its cytostatic wasobservedfornon-adherentBJABlymphomacells(Burkitt-like activity towards HT-29 and MCF-7 cells are observed for the lymphoma) following 3h incubation periods with complexes Me2dppz complex 7 when compared to the dppz complex 6. 1(Fig.S2a),5(Fig.S3),6(Fig.S2b)and8intheconcentrationrange d Su p r p p z r ) is a in re gl i y n , a b c o ti t v h e 8 (IC ( 5 p 0 p > ¼ 1 H 00 OO mM C- ) dp to p w z) ar a d n s d th 9 e ( s p a p m ¼ e n c - a B n u c N e H r C c O el - l 5 ca e u 5 s 0 e m n M eg . li T g h ib is le i c n e d l i l c d a e te a s th th b a y t n t e h c e ro o si r s g . anorhodium(III) complexes lines,astheyalsoaretowardsHEK-293cells.Thissuggeststhatthe The in vitro inhibition of BJAB cell proliferation bycomplexes presence of electron-withdrawing substituents mayhave a nega- 1,5,6and8wasalsoevaluated.Afteranincubationperiodof24h, tiveeffectonthecytotoxicityofthesecomplexes.Wehadconsid- theviabilityandcellcountweremeasuredwithaCASYCellCounter ered introducing a biotin-reporter moiety [36] at the end of andAnalyzerSystem,withthesettingsspecificallydefinedforthe a polyalkane spacer attached to the dppz ligand but have now requirementsoftheBJABcells.Thedose-dependentdecreaseincell abandonedthisprojectowingtothelackofactivityofcomplex9.It is interesting to note that the IC50 value of 2.80(cid:2)0.10mM for 1 pr a o n li d fe 8 ra in tio F n ig. is S4 d . e IC p 5 ic 0 t v e a d lu fo es r a c r o e m li p st le e x d e i s n 5 Ta a b n le d 4 6 .It in is F a i p g p . a 8 re a n n t d th fo a r t complexes 1, 5 and 6 are all effective at intermediate dose levels Table4 (IC50 ¼9e18mM).Althoughinactivetowardstheadherentcelllines MCF-7andHT-29,HOOC-dppzcomplex8doesinhibitproliferation Biologicaldataforcomplexes1,5,6and8inBJABcells. ofthenon-adherentBJABcells,albeitatasignificantlyhigherdose Compound pp IC50 a(mM) AC50 b(mM) level(IC50 ¼31mM). 1 Phen 11 21 Apoptosis,incontrasttounspecificnecrosis,requiresacontrolled 5 5,6-Me2phen 9 17 andregulatedmechanismleadingtocelldeath.DNAfragmentation 6 dppz 18 19 8 HOOC-dppz 31 >50 (hypoploidy) is considered to be a typical effect of apoptotic cell death,andwethereforequantifiedtheinductionofapoptosisfor1,5, a Inhibitionofproliferationafter24h. b Apoptosis induction after 72h; the AC50 values give concentrations of the 6and8byflowcytometricmeasurementsofDNAfragmentsafter complexesforwhich50%ofthecellsareundergoingapoptosis. incubatingBJABcellswiththecomplexesfor72h.Thenumbersof Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 1029 apoptotic cells for different concentrations of the complexes are RhCl}2(m-Cl)2] [17], [(h5-C5Me5)RhCl(phen)](CF3SO3) (1) [7], [(h5- illustratedinFig.9for5and6andinFig.S5for1. C5Me5)RhCl(dppz)](CF3SO3)(6)[15]andthedipyrido[3,2-a:20,30-c] AC50valuesfor50%apoptosisinductionafter72harelistedin phenazine 7,8-dimethyl (Me2dppz) and 2-carboxylic acid (HOOC- Table4.NosignificantDNAfragmentationwasobservedforBJAB dppz) derivatives [38,18] were synthesized according to literature cells after a 72h incubation period with concentrations of the procedures. HOOC-dppzcomplex8intherange0.1e25mM,andonly4.1%after treatment with 8 at a 50mM concentration. As can be seen from 4.1. Preparationof2e5and7e9 Fig.9andFig.S5,complexes1and5inducesignificantDNAfrag- mentation levels of, respectively, 14.3 and 28.3% at a 5mM 4.1.1. [(h5-C5Me5)RhCl(4-Mephen)](CF3SO3)(2) concentration. 38.8% of the BJAB cells are apoptotic following TwoequivalentsofAg(CF3SO3)(77.1mg,0.3mmol)wereadded treatmentwiththedppz6complexata10mMconcentration. to [{(h5-C5Me5)RhCl}2(m-Cl)2] (92.7mg, 0.15mmol) in 10ml acetone and stirred in the dark for 2h. Centrifugation of the 3. Conclusions resulting AgCl precipitate and subsequent solvent removal under vacuum afforded [(h5-C5Me5)RhCl(acetone)2](CF3SO3), which was Methyl substitution of phenanthroline and dppz ligands in stirredwith4-Mephen(58.3mg,0.3mmol)in10mlCH3OH/CH2Cl2 organorhodium(III) complexes of the type [(h5-C5Me5)RhCl(pp)] (1:1)for5hat75(cid:4)C.Theproductwassubsequentlyprecipitatedby (CF3SO3)leadstoasignificantincreaseincytostaticactivitytowards addition of diethyl ether, washed and dried in vacuo. Yield: 69% thehumancancercelllinesMCF-7andHT-29.Forphenanthroline (128mg).C24H25ClF3N2O3RhS:M¼616.9g/mol.Anal.Found(%):C, derivatives, increases in the antiproliferative properties are more 46.3;H,4.4;N,4.9;Calc.(%):C,46.7;H,4.1;N,4.5.LSIMS:m/z467 pronouncedforthe4-Mephenand4,7-Me2phencomplexes2and4 (100%) [M(cid:3)CF3SO3] þ , 431 (70%) [M(cid:3)CF3SO3 (cid:3)HCl] þ . 1H NMR than for the 5-Mephen and 5,6-Me2phen complexes 3 and 5. (CD2Cl2,200MHz,30(cid:4)C):d¼1.70(s,15H,Cp*eCH3),2.90(s,3H,4- Respectively 1.6- and 4.4-fold improvements in activity towards CH3),7.87(d,1H,H3),8.03(dd,1H,H8),8.09(d,1H,H5),8.17(d,1H, MCF-7andHT-29cellswereobservedfortheMe2dppzcomplex7 H6),8.62(d,1H,H7),8.95(d,1H,H2),9.12(d,1H,H9). incomparisontothedppzcomplex6.Incontrast,theHOOC-dppz and n-BuNHCO-dppz complexes 8 and 9 are inactive (IC50 > 4.1.2. [(h5-C5Me5)RhCl(5-Mephen)](CF3SO3)(3) 100mM)towardsboththeadherentcancercelllinesandadherent Preparation as for 2 with the ligand 5-Mephen (58.3mg, immortalized HEK-293 cells. Whereas electron-donating methyl 0.3mmol). Yield: 74% (136mg). C24H25ClF3N2O3RhS: M¼616.9g/ substituents cause an increase in cytotoxicity for 2e5 and 7, the mol.Anal.Found(%):C,47.1;H,4.1;N,4.9;Calc.(%):C,46.7;H,4.1;N, presence of electron-withdrawing groups in complexes 8 and 9 4.5. LSIMS: m/z 467 (100%) [M(cid:3)CF3SO3] þ , 431 (60%) leadstodrasticlossinactivity.InducednegativeCDbandsatabout [M(cid:3)CF3SO3 (cid:3)HCl] þ .1HNMR(CD2Cl2,200MHz,30(cid:4)C):d¼1.84(s, 296nmandlargeviscosityincreasesareinaccordancewithstable 15H,Cp*eCH3),2.93(s,3H,5-CH3),7.97(d,1H,H6),8.19(2dd,2H,H3/ DNA intercalation bycomplexes 7e9 with their substituted dppz 8),8.64(d,1H,H7),8.84(d,1H,H4),9.17(d,1H,H9),9.26(d,1H,H2). ligands. The lack of cytotoxic activity for 8 and 9, despite their ability to intercalate, suggests that DNA may well not be an 4.1.3. [(h5-C5Me5)RhCl(4,7-Me2phen)](CF3SO3)(4) importantintracellulartargetforthesehalf-sandwichrhodium(III) Preparation as for 2 with the ligand 4,7-Me2phen (62.5mg, complexes. This is most likely also the case for the Menphen 0.3mmol).Yield:75%(143mg).C25H27ClF3N2O3RhS:M¼630.9g/ complexes 2e5 (n¼1,2), for which noexperimental evidence for mol.Anal.Found(%):C,47.9;H,4.0;N,4.3;Calc.(%):C,47.6;H,4.3; DNA intercalation was obtained. It is possible that improved N, 4.4. LSIMS: m/z 481 (95%) [M(cid:3)CF3SO3] þ , 445 (100%) cellular uptake, increased electron density in the aromatic ring [M(cid:3)CF3SO3 (cid:3)HCl] þ .1HNMR(CD2Cl2,200MHz,30(cid:4)C):d¼1.84(s, systems or specific hydrophobic interactions in target binding 15H,Cp*eCH3),2.93(s,6H,4-CH3/7-CH3),7.85(d,2H,H3/8),8.18(s, pockets may lead to the observed increases in cytotoxicity on 2H,H5/6),8.94(d,2H,H2/9). methylsubstitutionofboththephenanddppzligands. Thenatureofcelldeathforlymphomacells(BJAB)wasstudied 4.1.4. [(h5-C5Me5)RhCl(5,6-Me2phen)](CF3SO3)(5) forcomplexes1,5and6withpp¼phen,5,6-Me2phenanddppz.All Preparation as for 2 with the ligand 5,6-Me2phen (62.5mg, three complexes induce significant apoptosis via the intrinsic 0.3mmol).Yield:65%(124mg).C25H27ClF3N2O3RhS:M¼630.9g/ mitochondrial pathway. The potential of these (h5-C5Me5)Rh(III) mol.Anal.Found(%):C,47.5;H,4.5;N,4.4;Calc.(%):C,47.6;H,4.3; polypyridylcomplexesisfurtherunderlinedbythefactthatthey N, 4.4. LSIMS: m/z 481 (87%) [M(cid:3)CF3SO3] þ , 445 (100%) causeonlynegligiblecelldeathbynecrosis. [M(cid:3)CF3SO3 (cid:3)HCl] þ .1HNMR(CD2Cl2,200MHz,30(cid:4)C):d¼1.69(s, 15H,Cp*eCH3),2.74(s,6H,5-CH3/6-CH3),8.05(dd,2H,H3/8),8.74 4. Experimental (d,2H,H4/7),9.05(d,2H,H2/9). Solvents were dried and distilled before use. An Analytik Jena 4.1.5. 4.1.5[(h5-C5Me5)RhCl(Me2dppz)](CF3SO3)(7) SPECORD200spectrometerwasemployedforUV/visiblemeasure- Preparation as for 2 with the ligand Me2dppz (93.1mg, mentsandCDspectrawereregisteredonaJascoJ-715instrument. 0.3mmol). Yield: 35% (77mg). C31H29ClF3N4O3RhS: M¼733.0g/ LSIMS(LiquidSecondaryIonmassSpectrometry)datawereregis- mol.Anal.Found(%):C,51.3;H,4.0;N,7.5;Calc.(%):C,50.8;H,4.0; teredwithaFisonsVGAutospecinstrumentemployingacesiumion N, 7.6. LSIMS: m/z 697 (10%) [M(cid:3)Cl] þ , 583 (96%) [M(cid:3)CF3SO3] þ , gun(voltage17kV)and3-nitrobenzylalcoholastheliquidmatrix. 548(100%)[M(cid:3)HCl(cid:3)CF3SO3] þ .1HNMR(CD2Cl2,200MHz,30(cid:4)C): ElementalanalyseswereperformedonaVarioEl(ElementarAna- d¼1.88(s,15H,Cp*eCH3),2.72(s,6H,dppz-CH3),8.26(s,2H,H11/ lysensysteme). RhCl3 $3H2O and Ag(CF3SO3) were purchased from 14),8.32(2dd,2H,H3/8),9.29(d,2H,H2/9),9.92(d,2H,H4/7). Chempur,1,10-phenanthroline(phen)andits4-methyl(4-Mephen), 5-methyl (5-Mephen), 4,7-dimethyl (4,7-Me2phen) and 5,6- 4.1.6. [(h5-C5Me5)RhCl(HOOC-dppz)](CF3SO3)(8) dimethyl (5,6-Me2phen) derivatives together with 3,4-diamino Preparation as for 2 with the ligand HOOC-dppz (97.9mg, benzoic acid, 4,5-dimethyl-1,2-phenyl-diamine, n-butylamine and 0.3mmol). Yield: 31% (70mg). C30H25ClF3N4O3RhS: M¼749.0g/ 1,10-carbonyldiimidazole(CDI)fromAcros.Solventswerepurchased mol.Anal.Found(%):C,47.6;H,3.4;N,7.8;Calc.(%):C,48.1;H,3.4; from TJ Baker and DEUTERO GmbH. The complexes [{(h5-C5Me5) N, 7.5. LSIMS: m/z 599 (100%) [M(cid:3)CF3SO3] þ , 563 (67%) 1030 Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 [M(cid:3)CF3SO3 (cid:3)HCl] þ .1HNMR(CD2Cl2,400MHz,30(cid:4)C):d¼1.86(s, removeparticulatematerialpriortouse.Reducedviscositieshwere 15H,Cp*eCH3),8.31(m,2H,H3/8),8.54(2d,2H,H13/14),9.11(s,1H, calculated by literature methods [25] and plotted as (h/h 0)1/3 H11),9.27(2d,2H,H2/9),9.92(2d,2H,H4/7). (h 0 ¼reduced viscosity of the DNA solution in the absence of complex)againstrforrod-likeDNA(approximately600basepairs). 4.1.7. 2-(7-n-Butylamido)dipyrido[3,2-a:20,30-c]phenazine, (n-BuNHCO-dppz) 4.4. NMRspectroscopy HOOC-dppz(163.2mg,0.5mmol)and1,10-carbonyldiimidazole (97.3mg, 0.6mmol) were stirred in 20ml pyridine for 1.5h at BrukerDPX400andDPX200spectrometerswereemployedfor 80(cid:4)C.Followingadditionofn-butylamine(49.6ml,0.5mmol),the the 1H NMR spectroscopic characterization of new compounds solutionwasheatedforafurther 2hat80(cid:4)Candthenfor1hat withtheresidual1Hsignalofthedeuteratedsolventbeingusedfor 25(cid:4)C.75mldiethyletherwasaddedandthereactionmixturewas chemicalshiftcalculation.Thesplittingsofprotonresonancesare left to stand for 18h at 4(cid:4)C. After solvent removal the resulting defined as s¼singlet, d¼doublet, t¼triplet and m¼multiplet. precipitatewasdissolvedinCH2Cl2andseparatedfromunsoluble Thehexanucleotidespectraford(50-CGTCGG-30)withWATERGATE materialbyfiltration.Theproductwasthenprecipitatedbyaddi- solventsuppressionwererecordedat600.13MHzprotonfrequency tionofdiethyletherat4(cid:4)Candwashedanddriedinvacuo.Yield: and at 288K on a Bruker DRX 600 spectrometer equipped with 26%(50mg).C23H19N5O:M¼381.2g/mol.Anal.Found(%):C,72.6; apulsedfieldgradientandatripleresonanceprobehead[41e43]. H, 4.8; N,18.3; Calc. (%): C, 72.4; H, 5.0; N,18.4. LSIMS: m/z 404 Theone-dimensional1HNMRspectrawererecordedwithatime (34%)[MþNa] þ ,382(100%)[MþH] þ ,327(43%)[M(cid:3)C4H7] þ .1H domainof32kdatapointsandaspectralwidthof12,019.23Hz.The NMR(DMSO-d6,200MHz,30(cid:4)C):d¼0.89(t,3H,n-BueCH3),1.34, sweep width of the two-dimensional homonuclear spectra was 1.40(2dd,4H,n-Bu-Hc/Hd),2.20(m,2H,n-Bu-Hb),2.89(dd,1H,n- 12,019.23Hz in the direct and in the indirect 1H dimension. The Bu-Ha),7.97(2dd,2H,H3/8),8.32(d,1H,H13)8.51(d,1H,H14),8.77 free-inductiondecaywasacquiredfor340.9msandthedwelltime (s,1H,H11),9.24(2d,2H,H2/9),9.61(2d,2H,H4/7). wassetto41.6ms.HomonuclearandNOESYspectrawerezero-filled prior to Fourier transformation and sine apodization functions 4.1.8. [(h5-C5Me5)RhCl(n-BuNHCO-dppz)](CF3SO3)(9) wereappliedinbothdimensions.Allspectrawereprocessedwith Preparationasfor2withtheligandn-BuNHCO-dppz(114.4mg, NMR-PipeandanalyzedwithNRMView[44,45].Assignmentsand 0.3mmol). Yield: 20% (49mg). C34H34ClF3N5O4RhS: M¼804.1g/ datahandlingwereperformedusingNMRView[45]. mol.Anal.Found(%):C,50.5;H,3.9;N,8.7;Calc.(%):C,50.8;H,4.3; N,8.7.LSIMS:m/z804(17%)[M(cid:3)H] þ ,713(76%)[M(cid:3)HCl(cid:3)C4H7] þ , 4.5. Cellcultures 563(100%)[M(cid:3)HCl(cid:3)H(CF3SO3)(cid:3)C4H7] þ . MC7-7 breast cancer and HT-29 colon carcinoma cells were 4.2. X-raystructuralanalyses maintained in DMEM high glucose (PAA) supplemented with gentamycin(50mgl (cid:3)1)and10%(v/v)fetalcalfserum(FCS)at37(cid:4)C Crystals of compounds 3 and 5 suitable for X-ray structural under 5% CO2 and passaged twice a week according to standard analysis were obtained by slow evaporation of CH3OH/H2O solu- procedures.ForHEK-293cellsDulbecco’smodifiedeaglemedium tions.TheircrystalandrefinementdataaresummarizedinTable1. (Invitrogen) supplemented with 10% fetal bovine serum and Intensity datawere collected on an Oxford Diffraction Xcalibur 2 100units/mlpenicillinandstreptomycinwasemployedat37(cid:4)C/5% diffractometerwithaSapphire2CCDdetectorusinguscansandMo CO2.Thecellsweresplitandaliquotswereseededin35mmculture Karadiation(l¼0.71073 (cid:2) A).Semi-empiricalabsorptioncorrections dishesthreetimesaweek.BJAB(Burkitt-likelymphoma)cellswere were applied to the intensities in all cases. The structures were maintained at 37(cid:4)C in RPMI 1640 (GIBCO, Invitrogen) supple- solvedbydirectmethodswithSHELXSandrefinedagainstF2with mented with 10% heat-inactivated FCS, penicillin (100,000Ul (cid:3)1), SHELXL[39].Anisotropictemperaturefactorswereintroducedfor streptomycin(0.1gl (cid:3)1)andL-glutamine(0.56gl (cid:3)1).Thecellswere non-hydrogen atoms and protons were refined at geometrically subcultured every 3e4 days by dilution of the cells to a concen- calculatedpositionsasridingatoms. tration of 1(cid:5)105cells/ml. Twenty-four hours before the assay setup,cellswereculturedataconcentrationof3(cid:5)105cells/mlto 4.3. DNA-bindingstudies ascertainstandardizedgrowthconditions.Fortheapoptosisassays, the cells were then diluted to a concentration of 1(cid:5)105cells/ml ThethermaldenaturationtemperaturesTmof1:5complex/DNA immediatelybeforeadditionofthecomplexes. mixtures [DNA concentration¼M(base pairs)] were determined for 1e9 in a 10mM phosphate buffer at pH¼7.2. Melting curves 4.6. Cytotoxicitymeasurements wererecordedat1(cid:4) stepsforthewavelengthl¼260nmwithan AnalytikJenaSPECORD200spectrometerequippedwithaPeltier Theantiproliferativeeffectsofthecompoundsweredetermined temperaturecontroller.Tmvalueswerecalculatedbydetermining followingestablishedprocedures[46,47].Cellsweresuspendedin the midpoints of melting curves from the first order derivatives. cell culture medium (HT-29: 2850cells/ml, MCF-7: 105cells/ml, TheexperimentalDTmvaluesareestimatedtobeaccuratewithin HEK-293:2000cells/ml),and100mlaliquotsthereofwereplatedin (cid:2)1(cid:4)C. Concentrations of CT DNA ([DNA]¼M(base pairs)) were 96well plates andincubated at 37(cid:4)C: 5%CO2 for 48h(HT-29or determined spectrophotometrically using the molar extinction HEK-293) or 72h (MCF-7). Stock solutions of the compounds in coefficient3 260 ¼13,200M (cid:3)1cm (cid:3)1[40]. DMFwerefreshlypreparedanddilutedwithcellculturemediumto Viscositiesformixturesofcomplexes7e9withsonicatedDNA thedesiredconcentrations(finalDMFconcentration:0.1%v/v).The weredeterminedusingaCannon-Ubbelhodesemi-microdilution medium in the plates was replaced with medium containing the viscometer (Series No 75, Cannon Instruments Co) held at compoundsingradedconcentrations(sixreplicates).Afterfurther a constant temperature of 25(cid:4)C in a water bath. The viscometer incubationfor72h(HT-29)or96h(MCF-7andHEK-293)thecell initially contained 2ml of 0.4mM sonicated DNA solution in biomass was determined by crystal violet staining and the IC50 a 10mM phosphate buffer (pH¼7.2). 0.2mM complex solutions valueswereestablishedasthoseconcentrationscausing50%inhi- alsocontaining0.4mMsonicatedDNAwereaddedinincrementsof bition of cell proliferation. Results were calculated from 2 to 3 100mlfromamicropipette.Solutionswerepassedthroughfiltersto independentexperiments. Y.Geldmacheretal./JournalofOrganometallicChemistry696(2011)1023e1031 1031 Thecytotoxicityofcomplexes1,5,6and8towardsBJABcellswas References investigatedbyreleaseoflactatedehydrogenase(LDH)asdescribed previously[37].Afterincubationwithdifferentconcentrationsofthe [1] W.S.Sheldrick,S.Heeb,Inorg.Chim.Acta168(1990)93e100. complexesfor1hor3hat37(cid:4)C,LDHactivityreleasedbyBJABcells [2] Y 47 .W 64 . e Y 4 a 7 n 7 , 6 M . .Melchart,A.Habtemarian,P.J.Sadler,Chem.Commun.(2005) wasmeasuredinthecellculturesupernatantsusingtheCytotoxicity [3] S. Schäfer, I. Ott, R. Gust, W.S. Sheldrick, Eur. J. Inorg. Chem. (2007) DetectionKitfromBoehringerMannheim (cid:3) (Mannheim,Germany). 3034e3046. [4] P.J.Dyson,G.Save,DaltonTrans.(2006)1929e1933. Thesupernatantswerecentrifugedat1500rpmfor5min.Cell-free [5] A.Dorcier,W.H.Ang,S.Bolan(cid:3)o,L.Gonsalvi,L.Juillerat-Jeannerat,G.Laurenczy, supernatants(20ml)weredilutedwithphosphate-bufferedsaline M.Peruzzini,A.D.Phillips,F.Zanobini,P.J.Dyson,Organometallics25(2006) (PBS, 80ml) and a reaction mixture containing 2-[4-iodophenyl]- 4090e4096. [6] S.Schäfer,W.S.Sheldrick,J.Organomet.Chem.692(2007)1300e1309. 3-[4-nitrophenyl]-5-phenyltetrazolium chloride (INT), sodium þ [7] M.A.Scharwitz,I.Ott,Y.Geldmacher,R.Gust,W.S.Sheldrick,J.Organomet. lactate, oxidised nicotinamide adenine dinucleotide (NAD ) and Chem.693(2008)2299e2309. diaphorase(100ml)wasadded.Time-dependentformationofthe [8] M.Gras,B.Therrien,G.Suss-Funk,A.Casini,F.Edafe,P.J.Dyson,J.Organomet. reactionproductwasthequantifiedphotometricallyat490nm.The Chem.695(2010)1119e1125. [9] S. Wirth, C.J. Rohbogner, M. Cieslak, J. Kazmierczak-Baranska, S. Donevski, maximumamountofLDHactivityreleasedbythecellswasdeter- B.Nawrot,I.P.Lorenz,J.Biol.Inorg.Chem.15(2010)429e440. mined by lysis of the cells by using 0.1% Triton X-100 in culture [10] R. Bieda, I. Ott, M. Dobrosche, A. Prokop, R. Gust, W.S. Sheldrick, J. Inorg. mediumandsetas100%celldeath. Biochem.103(2009)698e708. (cid:3) [11] R.Bieda,I.Ott,R.Gust,W.S.Sheldrick,Eur.J.Inorg.Chem.(2009)3821e3831. The viability was determined by using the CASY Cell Coun- [12] W.D. McFadyen, P.G.W. Lawrence, I.A.G. Roos, V.A. Leopold, J. Med. Chem. terþAnalyzerSystemfromInnovatis(Bielefeld,Germany).Settings 28(1985)1113e1116. were specifically defined for the requirements of the used cells. [13] C.R.Brodie,J.G.Collins,J.R.Aldrich-Wright,DaltonTrans.(2004)1145e1152. [14] A.M.Krause-Heuer,R.Grünert,S.Kühne,M.Buczkowski,N.J.Wheate,D.D.Le Withthissystem,thecellconcentrationcanbeanalyzedsimulta- Pevelen, L.R. Boag, D.M. Fisher, J. Kasparkova, J. Malina, P.J. Bednarski, neously in three differentsize ranges: cell debris, dead cells, and V.Brabec,J.R.Aldrich-Wright,J.Med.Chem.52(2009)5474e5484. viable cells. Cells were seeded at a density of 105cells/ml and [15] D.A.Herebian,W.S.Sheldrick,J.Chem.Soc.,DaltonTrans.(2002)966e974. [16] M.Dobrosche,Y.Geldmacher,I.Ott,M.Harlos,L.Kater,L.Wagner,R.Gust, treated with different concentrations of complexes 1, 5, 6 and 8; W.S.Sheldrick,A.Prokop,ChemMedChem4(2009)177e187. non-treatedcellsservedascontrols.Aftera24hincubationperiod [17] J.W.Kang,K.Moseley,P.M.Maitlis,J.Am.Chem.Soc.91(1969)5970e5977. at37(cid:4)C,cellswereresuspendedproperlyand100mLofeachwell [18] B. Gholamkhass, K. Koike, N. Negishi, H. Hori, K. Takeuchi, Inorg. Chem. 40(2001)756e765. was diluted in 10ml CASYton (ready-to-use isotonic saline solu- [19] L. Dadci, H. Elias, U. Frey, A. Hörnig, U. Koelle, A.F. Merbach, H. Paulus, tion)foranimmediateautomatedcountofthecells. J.S.Schneider,Inorg.Chem.34(1995)306e315. [20] M.-T.Youinou,R.Ziessel,J.Organomet.Chem.363(1989)197e208. [21] W.C.Johnson,in:N.Berova,K.Nakanishi,R.W.Woody(Eds.),CircularDichroism: 4.7. MeasurementofDNAfragmentation PrinciplesandApplications,seconded.VCHPublishers,NewYork,2000,p.523. [22] X.-N.Li,Z.-J.Wu,Z.-J.Si,L.-Z.Liu,X.-J.Liu,H.-J.Zhang,Phys.Chem.Chem. Phys.11(2009)9687e9695. Apoptoticcelldeath was determined bya modifiedcell cycle- [23] J. Olofsson, L.M. Wilhelmsson, P. Lincoln, J. Am. Chem. Soc. 126 (2004) analysis,whichdetectsDNAfragmentationatthesinglecelllevel 15458e15465. [48,49].Cellswereseededatadensityof105cells/mlandtreated [24] K. O’Donoghue, J.C. Penedo, J.M. Kelly, P.E. Kruger, Dalton Trans. (2005) 1123e1128. withdifferentconcentrationsof1,5,6and8.Aftera72hincubation [25] G.Cohen,H.Eisenberg,Biopolymers4(1966)429e440. periodat37(cid:4)C,cellswerecollectedbycentrifugationat1500rpm [26] D.Suh,J.B.Chares,Bioorg.Med.Chem.3(1995)723e728. for 5min, washed with PBS at 4(cid:4)C and fixed in PBS/2% (v/v) [27] M.E.Hogan,N.Dattagupta,D.M.Crowthers,Biochemistry18(1979)280e288. formaldehydeonicefor30min.Afterfixation,cellswerepelleted, [28] N M . e J. d W .C h h e e a m te . , 7 C. ( R 2 . 0 B 0 r 7 o ) d 6 ie 2 , 7 J e .G 6 . 4 C 8 o . llins,S.Kemp,J.R.Aldrich-Wright,Mini-Rev. incubated with ethanol/PBS (2:1, v/v) for 15min, pelleted for [29] C.M.Dupureur,J.K.Barton,J.Am.Chem.Soc.116(1994)10286e10287. asecondtimeandresuspendedinPBScontaining40mgml (cid:3)1RNase [30] J.G.Collins,A.D.Steedmann,J.R.Aldrich-Wright,I.Greguric,T.W.Hambley, A. RNAwas digestedfor 30min 37(cid:4)C,after which the cells were Inorg.Chem.37(1998)3133e3141. [31] Mhd.A.Nazif, J.-A.Bangert,I.Ott, R. Gust,R.Stoll, W.S.Sheldrick, J.Inorg. pelleted once again and finally resuspended in PBS containing Biochem.103(2009)1405e1414. 50mgml (cid:3)1 propidium iodide. Nuclear DNA fragmentation was [32] J. Feigon, W.A. Denny, W. Leupin, D.R. Kearns, J. Med. Chem. 27 (1984) quantified by flowcytometric determination of hypodiploid DNA [33] 4 M 5 . 0 K e o 4 c 6 h 5 o . yan,J.L.Leroy,M.Gueron,J.Mol.Biol.196(1987)599e609. (Fluorescence-activated cell sort, FACS). Data were collected and [34] A.Myari,N.Hadjiiliandis,A.Garoufis,J.Malina,V.Brabec,J.Biol.Inorg.Chem. analyzedusingaFACScan(BectonDickinson,Heidelberg,Germany) 12(2007)279e292. [35] A. Frodl, D. Herebian, W.S. Sheldrick, J. Chem. Soc., Dalton Trans. (2002) apparatusequippedwiththeCELLQuestsoftware.Dataaregivenin 3664e3673. %hypoploidy(subG1),whichreflectsthenumberofapoptoticcells. [36] K.K.-W.Lo,K.H.-K.Tsang,K.-S.Sze,Inorg.Chem.45(2006)1714e1722. [37] R.A.Diller,H.M.Riepl,O.Rose,C.Frias,G.Henze,A.Prokop,Chem.Biodivers. 2(2005)1331e1337. Acknowledgements [38] R.E.Holmlin,J.A.Yao,J.K.Barton,Inorg.Chem.38(1999)174e189. [39] G.M.Sheldrick,SHELXS97andSHELXL97.UniversityofGöttingen,Germany, Financial support of this work in Bochum, Braunschweig and 1997. [40] J.Marmur,J.Mol.Biol.3(1961)208e218. Köln by the Deutsche Forschungsgemeinschaft (DFG) within the [41] L.Braunschweiler,R.R.Ernst,J.Magn.Reson.53(1983)521e528. research group FOR 630 “Biological function of Organometallic [42] D.G.Davis,A.Bax,J.Am.Chem.Soc.107(1985)7197e7198. Compounds”isgratefullyacknowledged. [43] A.J.Shako,C.J.Lee,A.Pines,J.Magn.Reson.77(1988)274e293. [44] F.Delagio,S.Grzesiek,G.W.Vuister,G.Zhu,J.Pfefer,A.Bax,J.Biomol.NMR 6(1995)277e293. [45] R.D.Johnson,P.Bluemler,R.Rafey,D.Brodbeck,Abstr.Pap.Am.Chem.Soc. AppendixA. Supplementarymaterial 207(1994)138-COMP. [46] I.Ott,K.Schmidt,B.Kircher,P.Schuhmacher,T.Wiglenda,R.Gust,J.Med. CCDC 782259 and 782260 contain the supplementarycrystal- Chem.48(2005)622e629. [47] R. Bieda, M. Dobroschke, A. Triller, I. Ott, M. Spehr, R. Gust, A. Prokop, lographic data for this paper. These data can be obtained free of W.S.Sheldrick,ChemMedChem5(2010)1123e1133. chargefromTheCambridgeCrystallographicDataCenterviawww. [48] P.K. Smith, R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H. Gartner, cccdc.cam.ac.uk/data-request/cif. M.D. Provenzano, E.K. Fujimoto, N.M. Goeke, B.J. Olson, D.C. Klenk, Anal. Biochem.150(1985)76e85. Supplementarymaterialassociatedwiththisarticlecanbefound, [49] T. Wieder, A. Prokop, B. Bagci, F. Essmann, D. Bernicke, K. Schulze-Osthoff, intheonlineversion,atdoi:10.1016/j.jorganchem.2010.10.034. B.Dörken,H.-G.Schmalz,P.T.Daniel,G.Henze,Leukemia15(2001)1735e1742.