Molecular interactions of ruthenium complexes in isolated mammalian nuclei and cytotoxicity on V79 cells in culture
Post on 15-Sep-2016
.Mutation Research 423 1999 171181
Molecular interactions of ruthenium complexes in isolatedmammalian nuclei and cytotoxicity on V79 cells in culture
Andrea Barca a, Bianca Pani a, Marisa Tamaro b, Elio Russo a,)a Dipartimento di Biochimica, Biofisica e Chimica delle Macromolecole, Uniersita di Trieste, Via L. Giogieri 1, 34127 Trieste, Italy`
b Dipartimento di Scienze Biomediche, Uniersita di Trieste, Trieste, Italy`
Received 15 July 1998; revised 17 November 1998; accepted 18 November 1998
In this paper, the molecular interactions in isolated mammalian nuclei of three ruthenium complexes, which are putativeantineoplastic chemotherapeutic agents effective in reducing metastatic tumours in vivo, have been investigated and
.compared with the well-known antitumour drug CDDP cis-diamminedichloroplatinum . The compounds studied are: . . . . .Na trans-RuCl DMSO Imidazole NAMI , Na trans-RuCl DMSO Oxazole NAOX and Na trans-RuCl TMSO -4 4 4
.Isoquinoline TEQU . This study shows that the drugs bind to DNA but induce few, if any, DNA interstrand crosslinks,which are considered as the main biological lesions involved in the cytotoxic activity of several already known antitumourdrugs, whilst in the same experimental conditions, CDDP is confirmed to induce them. On the other hand, proteins appear tobe an important target in the cell for these drugs, since proteins-DNA crosslinks are shown to be induced by the complexes.Moreover, we investigated Ru complexes for their direct cytotoxicity on V79 cells in culture, showing that two of them .NAMI and NAOX do not significantly reduce the cloning efficiency of the cells even at concentrations as high as 23mgrml: only TEQU both reduces cloning efficiency and induces a significant number of mutants in V79 cells in culture.q 1999 Elsevier Science B.V. All rights reserved.
.Keywords: Cytotoxicity; DNADNA crosslink; DNAprotein crosslink; Ruthenium III complex
Heavy metals coordination complexes such asplatinum, gold, ruthenium, etc. have been investi-
Abbreviations: CDDP, cis-Diamminedichloroplatinum; NAMI, .N a trans-RuCl D M SO im idazole; N A O X , N a trans-4
. .RuCl DMSO oxazole; TEQU, Na trans-RuCl TMSO isoquino-4 4line; CFC, Colony forming cells; TMSO, Tetramethilenesulfoxide;FCS, Fetal calf serum; DMEM, Dulbeccos minimal essentialmedium; 6-TG, 6-Thioguanine; HGPRT, Hypoxanthineguaninephosphoribosyl transferase
) Corresponding author. Tel.: q39-040-676-3679; Fax: q39-040-676-3691; E-mail: email@example.com
gated for their possible antitumour activity. Amongthe most studied metal complexes, platinum deriva-tives have been shown to be the most promisingchemotherapeutic agents against mammalian tu-mours: at present, CDDP has proven to be veryeffective in clinical therapy of several human solidtumours such as testicular carcinomas, ovarian tu-mours, head and neck cancers, bladder tumours andosteosarcomas; however, it shows only a weak effectagainst many other malignancies of relevant socialincidence such as breast cancers, lung and colorectal
w xadenocarcinomas 1,2 . For these reasons, new metalcoordination complexes have been studied in order
0027-5107r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. .PII: S0027-5107 98 00240-1
( )A. Barca et al.rMutation Research 423 1999 171181172
to find compounds active against tumours that do notrespond, or show resistance, to CDDP and with lesstoxic side effects. In particular, some rutheniumderivatives have been in recent years synthetized andscreened for their antineoplastic activity and for their
w xcytotoxic effects 3 . Two complexes, namelyw . x w . xcis RuCl NH Cl and fac RuCl NH have2 3 4 3 3 3
been shown to be active against P388 mouse . .leukemia. Since then, many other Ru III and Ru II
derivatives have been synthetized and studied, withdifferent ligands like heterocyclic nitrogen ligands .imidazole or indazole or sulfoxides. The release ofchloride ligands should allow the interaction withbiological targets by formation of covalent bonds; sothe nature of ligands affects the biological activity ofruthenium compounds. A mechanism of activationby reduction was claimed to explain their activity:
.according to this hypothesis Ru III complexes canbe considered as prodrugs which should be activated
.by reduction to the corresponding Ru II specieswhich in turn should act as the true biological
w xreagents 4,5 . It should be noted that in solid tumourtissues the environment is considered to be hypoxic,
.so that the reduction of Ru III to the active species .Ru II is facilitated and at the same time, its reoxida-
tion becomes unlikely. This fact could induce an .accumulation of active species of Ru II compounds
just inside the solid tumour tissues, hence, makingthe cytotoxicity of these molecules against tumoursselective with respect to the normal tissues. More-over, because of the similarities between iron andruthenium, the latter seems to enter the cells throughthe Fe-transferrin system: since this transport proteinis more expressed in rapidly growing cells whichshow an increased iron requirement, ruthenium accu-
w xmulates into neoplastic cells 5 .DNA is generally considered the main target for
antineoplastic drugs acting as alkylating drugs likew xmetallic complexes 1,68 , but probably other
molecular interactions are relevant for their biologi-cal effects.
In this paper, three Ru complexes, namely: Na- . .trans-RuCl DMSO Imidazole NAMI , Na trans-4
. .RuCl DMSO Oxazole NAOX , and Na trans-4 . .RuCl TMSO Isoquinoline TEQU , whose formulas4
are shown in Fig. 1, have been investigated for theirinteractions with nuclear chromatin, looking for theformation of DNADNA and DNAproteins
Fig. 1. Chemical formulas of ruthenium complexes investigated.Abbreviations: Ims imidazole; Iqs Isoquinoline; Ox sOxazole;DMSOsDimethylsulfoxide; TMSOsTetramethylenesulfoxide.
.crosslinks. This series of ruthenium III complexes,structurally related and characterized by the presenceof sulfoxide and nitrogen-donor ligands, were testedon TLX5 lymphoma and some of them on MCamammary carcinoma in order to evaluate the rela-tionship of cytotoxicity and anti-metastatic activitywith their respective chemical properties. The drugswere shown to be cytotoxic only at high concentra-
y4 .tions )10 M and their cytotoxicity is related tolipophilicity. The comparison of the in vitro cytotox-icity and in vivo antitumour and antimetastatic activ-ity showed that the reduction of metastasis formationis not related to a direct cytotoxicity on tumour cells.In particular, the most cytotoxic compound, TEQU,is the least effective in reducing metastases, whilstNAMI which is very effective in reducing metastasesformation is slightly cytotoxic on tumour cells in
w xvitro 9 .In this paper, the cytotoxic activity of the drugs
has been also studied on mammalian cells in cultureand the mutagenic activity of the most cytotoxic onehas been evaluated in the V79rHGPRT system.
2. Materials and methods
CDDP, DMEM and DNase I were purchasedfrom Sigma, pBR322, HindIII, RNase and Pro-
( )A. Barca et al.rMutation Research 423 1999 171181 173
teinase K from Boehringer Mannheim, micrococcalnuclease from Amersham International, trypsin fromDifco Laboratories, FCS from BioSpa. Rutheniumcomplexes were synthetized as described and kindly
supplied by Prof. Mestroni Dipartimento di Scienze. w xChimiche, University of Trieste, Italy 10 . The
drugs were dissolved in water just before use at theappropriate concentrations. All other chemicals werepurchased from chemical sources.
2.2. Reactions with complexes
pBR322 Plasmid was linearized with HindIII,w xaccording to standard protocols 11 , and purified by
spinning 20 min at 500 g on Microcon 100 Amicon.Grace , changing buffer to 40 mM Tris pH 7.5, 1
.mM EDTA. 100 ng DNA supercoiled or linear atthe concentration of 25 ngrml were reacted for 1 hat 378C with metal complexes at the desired molarratio. The reactions were quenched with 1.5 MAcONa and 100 mM EDTA, DNA was collected byprecipitation with ethanol, dissolved in 40 mMTrisAcONa pH 8.2, 1 mM EDTA, 10% glyceroland analysed by 1% agarose gel electrophoresis inthe presence of 0.5 mgrml Ethidium Bromide for 1h at 50 V. Samples for denaturation experimentswere heated 2 min at 908C in the presence of 30%DMSO and immediately cooled on ice just before
w xloading 12 . The following 30-bp double-strandedoligonucleotide was used for footprinting experi-ments:
CGCThe two strands were synthetized separately by
solid-phase procedures using standard phospho-roamidite chemistry Applied Biosystem model 380
.B DNA synthetizer . The fully deprotected oligomerswere purified by anion-exchange chromatography
.using a Mono-Q HR column Pharmacia and even-w 32 xtually labeled with g- P ATP and T4 polynu-
.cleotide kinase Pharmacia . The two strands weremixed at the right stoichiometry to obtain the doublehelix oligonucleotides with only one strand labeled.A total of 0.4 pmoles were reacted 12 h at 378C with
.an excess of Ru complex 10 pmoles in 10 ml Tris50 mMNaCl 10 mM. Then the solutions were made
10 mM MgCl and treated with 1 ng of DNase I at2378C. Aliquots were taken at different times,quenched with 100 mM EDTA and analyzed byelectrophoresis in 20% denaturing polyacrylamidegels in the presence of 8 M urea at 458C. Followingelectrophoresis, the gels were transferred onto What-man 3MM chromatographic paper, dried and ex-posed to autoradiography for about 8 h.
w xNuclei from rat thymus prepared as described 13and stored in the presence of 50% glycerol at y208C
6 were treated as follows: 80=10 nuclei ;500 mg.DNA were incubated for 8 h at 378C with the
complexes at different concentrations expressed as. molrbp , in 500 ml RSB 10 mM TrisHCl pH 7.4,
.10 mM NaCl, 5 mM MgCl . The reactions were2stopped by cooling samples on ice, centrifuging at1000=g and resuspending in 500 ml cold RSB.
For the digestion with micrococcal nuclease, sam- .ples ;50 mg DNA each were suspended in 50 ml
10 mM TrisHCl pH 7.4, 1 mM CaCl and digested2at 378C with 25 units of micrococcal nuclease. Fiveminutes incubation were sufficient in order to obtainmainly mono- and a little amount of oligo-nucleo-somes in untreated nuclei. The reactions werequenched by adding 5 ml 5% SDS and 50 mMEDTA and cooling on ice. Samples were then depro-teinized by incubating twice for 3 h at 508C with 5mg of Proteinase K, adjusted to 1 M NaCl and
.extracted with chloroformrisoamyl alcohol 24:1 .DNA, collected by precipitation with ethanol, was
dissolved in 200 ml TBE 90 mM Tris, 80 mM.H BO , 2.5 mM EDTA pH 8 with 10% glycerol3 3
and analysed by 1.5% agarose gel electrophoresis inTBE in the presence of 0.5 mgrml Ethidium Bro-mide for 2 h at 70 V.
For the digestion with DNAse I, nuclei ;200.mg DNA were suspended in 200 ml 10 mM Tris
HCl pH 7.4, 5 mM EDTA, 0.5% SDS and digestedfor 1 h at 378C with 20 mg RNase, then twice for 3 hat 508C with 20 mg Proteinase K. Samples, adjustedto 1 M NaCl, were extracted with chloroformriso-
.amyl alcohol 24:1 and DNA was collected byprecipitation with ethanol. A total of 25 mg purifiedDNA were digested at 208C with 0.5 units of DNaseI in 1 ml 10 mM TrisHCl pH 7.4, 2 mM MnCl .2The reaction was followed by a Jasco V550 spectro-photometer at 260 nm in a 1 cm quartz cell. Datawere collected starting 30 s after the addition of the
( )A. Barca et al.rMutation Research 423 1999 171181174
enzyme and plotted as absorbance increase againsttime, showing a linear trend and allowing the calcu-lation of initial digestion rate. The percent inhibition .I of digestion rate is expressed by the equation:Is 1yr 100 .t cwhere and are initial DNA digestion rates fromt crespectively treated and control nuclei.
.Nuclei ;200 mg DNA each were extractedovernight with 400 ml 0.4 M H SO and free pro-2 4teins were determined in the supernatant through the
w xBradford method 14 .
2.3. DNA melting
.Untreated and treated nuclei ;50 mg DNA eachwere suspended in 50 ml 10 mM TrisHCl pH 7.4, 5mM EDTA, 0.5% SDS and digested for 1 h at 378Cwith 20 mg RNase, then twice for 3 h at 508C with20 mg Proteinase K. Samples, adjusted to 1 M NaCl,were extracted with chloroformrisoamyl alcohol .24:1 and DNA collected by pre-cipitation withethanol was redissolved in 1.5 ml of SDS 1%,transferred in quartz cells and the absorbance at 260nm was recorded by a Jasco V 550 spectrophotom-eter, changing the temperature by 18Crmin.
2.4. Cytotoxicity and mutagenicity assay on V79Chinese hamster cells
For the evaluation of cytotoxicity of Ru com-pounds, 22.5=106 V79 Chinese Hamster cellswere plated on 2000 mm2 Petri dishes in DMEMsupplemented with FCS, 100 UIrml penicillin and100 grml streptomycin. After 24 h incubation at378C in a humidified CO incubator, the medium2was replaced with fresh DMEM without FCS con-taining different concentrations of each Ru complexobtained by dilution of freshly prepared 10 mgrmlsolutions of each compound. All experiments in-cluded a control culture, in which the medium wasreplaced with fresh DMEM without FCS. After 1 hincubation, treated and untreated cultures werewashed three times with medium without FCS, andthen detached from the plates with trypsin. When
.necessary see Section 4 , the cells were detachedmechanically with a rubber policeman. The cells ofeach treated or untreated culture were then re-sus-
pended in DMEM with FCS, the suspensions werecounted in a Burker haemocytometer and replated at
2low density 200 cellsr2000 mm Petri dish, four.replicates .
The cloning efficiency was evaluated after 7 daysincubation by direct counting of more than 100 cellscolonies after staining with 0.1% methylene blue.The cloning efficiency is expressed as percent CFCwith respect to the control culture.
The mutagenesis assay was performed accordingw xto ONeill with minor modifications 15 . Together
with the cultures performed to evaluate the cloningefficiency, subcultures of 7.5=105 cells in 8000mm2 dishes were grown from each treated or un-
treated sample. Six to eight days later which is thepreviously determined appropriate expression time
.for HGPRT mutants , cells from each subculturewere detached with trypsin, re-suspended in com-
. 5plete medium and re-plated as follows: a 2=102 .cells per 8000 mm dish five replicates to which, 1
.h after plating, 6-TG 4 mgrml final concentration .was added in order to select HGPRT mutants. b
2 2 .2=10 cells per 2000 mm dish four replicates forthe evaluation of CFC.
After further 7 days incubation, colonies werescored by staining with methylene blue in both seriesof plates in order to evaluate the number of 6-TGresistant mutants per 106 CFC.
3.1. Reaction with naked DNA
We used linearized pBR322 DNA to investigatewhether the complexes are able to induce crosslinks
w xon DNA 12 . After treatment with the drugs at thedesired molrbp ratio, samples were purified by pre-cipitation with ethanol to remove the unreacted com-plexes and then thermally denaturated and renatu-rated just before loading on agarose gel. Denaturated
.linear single-stranded SS DNA is well resolved .from double stranded DNA DS since the former
migrates faster and is stained to a less extent byEthidium Bromide. Fig. 2 shows that, as previouslydetermined, in CDDP treated samples DS DNA isstill present after denaturation, whilst in Ru treatedsamples DS DNA is found only in TEQU.
( )A. Barca et al.rMutation Research 423 1999 171181 175
Fig. 2. Agarose gel electrophoresis of linearized pBR322 plasmid treated with drugs and thermally denaturated. Drug concentrations areexpressed as molrbp. DSsdouble strand DNA; SSsSingle strand DNA.
When a double-strand break is introduced into .DNA the supercoiled form S of the plasmid is .transformed in the linear one L , while in the case
of the formation of single-strand nicks an increase in .the amount of the relaxed form R is noted. These
three forms are easily separated by agarose elec-trophoresis in the presence of Ethidium Bromide.Fig. 3 shows the electrophoresis of circular pBR322plasmid treated with the drugs at different concentra-tions. It is evident that no compound examined isable to introduce double-strand breaks since L banddoes not appear in any case. The scanning of elec-trophoresis photos gives the ratio R formrtotal DNA.Without drugs, ;20% of plasmid is in the R bandand at a concentration of 0.5 molrbp that amountincreases by 28% with CDDP, by 1% with NAMI,by 2% with NAOX and by 4% with TEQU. At ourknowledge that is the first time that one demon-strates that CDDP introduces single strand nicks inDNA. Since the effect of the treatment with the Rucomplexes is very slight, we investigated Ru drugs athigher concentrations obtaining increments in the Rform of 4% for NAMI and 7% for NAOX at 2molrbp, and of 12% for TEQU at 1 molrbp, soconfirming the effects induced by these drugs.
Dnase I footprinting experiments were performed .on a 30-bp DNA composed by a pyrimidine PY
.and a purine PU strand, that contains respectively
Fig. 3. Agarose gel electrophoresis of supercoiled pBR322 plas-mid treated with drugs. Drug concentrations are expressed asmolrbp. Rs relaxed, Sssupercoiled, L s linear plasmid.
( )A. Barca et al.rMutation Research 423 1999 171181176
74% of CT or of GA bases. The DNA fragment wasalternatively labeled on one strand and the test wasdone to detect whether complexes bind to nuc-leotides and to stress preferences for bases. In Fig. 4,we show results obtained with TEQU. All treatedoligonucleotides are less digested than untreated ones.The effect for the PU-strand is not very strong, but itis a little more evident for the PY-strand. Theseresults suggest that the drug binds to DNA withoutaltering too much its conformation and that there ismore affinity towards pyrimidinic moieties.
3.2. Reaction with nuclei
The in vivo situation was approached using puri-fied nuclei to test complexes, extending incubationtime to 8 h, since preliminary tests revealed noappreciable variations yet after 4 h. Nuclei washed
and collected through centrifugation to remove theunreacted drugs appeared yellow stained, confirmingthat complexes had bound to chromatin. We havelimited maximum concentration range to 0.5 molrbpsince molarity of drugs was relatively elevated and athigher values, self-polymerization of Ru complexes
w xoccurred 16 . Actually with NAOX that happenedeven at 0.5 molrbp, so to compromise further han-dling of samples especially in preventing the com-plete extraction of different components; thereforedata obtained with NAOX at 0.5 molrbp are notwell comparable with those obtained with other com-plexes.
Micrococcal nuclease digestion is affected bymodifications in the higher order chromatin, since itsfirst target is linker DNA, without any sequencespecificity. In fact, Pt treated nuclei are more resis-
w xtant to the enzyme than untreated ones 17 . Fig. 5
Fig. 4. DNA footprinting with Dnase I. Lanes 1, 2, 5, 6, 9, 10, 13 and 14 contain 0.4 pmoles of the oligonucleotide with the PY strand5X-labelled. Lanes 3, 4, 7, 8, 11, 12, 15 and 16 contain 0.4 pmoles of the oligonucleotide with the PU strand 5X-labelled. The untreated DNAis in the odd lanes, whilst the treated DNA is in the pair lanes. Digestion times with DNA I: lanes 14, 30 s; lanes 58, 45 s; lanes 912, 60s; lanes 1316, 120 s.
( )A. Barca et al.rMutation Research 423 1999 171181 177
Fig. 5. Agarose gel electrophoresis of nuclei digested with micrococcal nuclease. Drug concentrations are expressed as molrbp. As0;Bs0.125; Cs0.25; Ds0.5. Msmono, Dsdi-, Ts tri- and Tes tetranucleosomes.
shows the results obtained when digestion times .were adjusted in order to obtain mainly mono- M
. .and di-nucleosomes D in control nuclei lanes A .The electrophoretic patterns of samples from treated
.nuclei lanes B, C and D show particles of increas- . .ing size like tri- T and tetra-nucleosomes Te , with
the characteristic 200 base pairs pattern. This effectis clearly concentration dependent and is very similarfor NAMI and TEQU, showing some difference forNAOX at the highest concentration, as above dis-cussed. These results reflect DNA distortions causedby the binding of the drugs to chromatin, as already
w xdemonstrated for CDDP 18 .The reduced activity of micrococcal nuclease in
treated samples could depend on modifications in-duced in chromatin bound proteins; therefore, afterthe treatment with the complexes, we have digestedall proteins with Proteinase K and have treated thepurified DNA with DNase I using as metallic cofac-tor Mnqq, to minimize differences due to the DNA
w xsequence 19 . The initial digestion rates plottedagainst drug concentration were linear from 30 untilto at least 250 s after the addition of the enzyme.Fig. 6 shows the reduction of initial digestion ratedue to the drugs. The effect is very low for TEQU,quite pronounced for NAOX, whilst NAMI andCDDP induce a comparable intermediate rate reduc-tion. CDDP data are consistent with previously re-
w xported results 20 .
3.3. DNA melting
The ability of Ru complexes to introduce DNAinterstrand crosslinks in nuclei was investigated bymelting experiments. Purified DNA was first heated
Fig. 6. Percent inhibition of the DNase I digestion of nucleitreated at different drugs concentrations.
( )A. Barca et al.rMutation Research 423 1999 171181178
.to 808C above melting temperature , and then cooledto 408C. Absorbances at 260 nm were recorded, andnormalized values were plotted against temperature .Fig. 7 . In these conditions, the renaturation ofmammalian genomic DNA is normally very limited,only 67% reverting to the native conformation, butthe presence of interstrand crosslinks favours theexact base pairing so promoting a rapid renaturationw x21 . The validity of determining the fraction of fastrenaturing DNA after denaturation to measurecrosslink amount has been demonstrated by Pera et
w xal. 22 who have compared this method with alka-line elution and DNA sedimentation obtaining simi-lar results. The spectrophotometric analysis showsalso the cooperative trend of the event. Whilst DNAfrom CDDP treated nuclei shows an amount of fastrenaturing DNA which increases with the drug con-
.centration curves B, C and D , DNA from nucleitreated with Ru complexes at 0.5 molrbp renaturesvery poorly, at an extent very similar for the three
.drugs curves E, F and G and for the untreated .sample curve A . These data are not completely in
.agreement with previously described results Fig. 2 ,where we demonstrated that TEQU was able tointroduce crosslinks in DNA.
Fig. 7. Renaturation profiles of DNA extracted from nuclei treatedat different drugs concentrations. Concentrations are expressed asmolrbp. Asuntreated DNA; Bs0.125, Cs0.25, Ds0.5 molCDDPrbp; Es0.5 mol NAMIrbp; Fs0.5 mol NAOXrbp;Gs0.5 mol TEQUrbp.
Fig. 8. Percent retention of proteins extracted from nuclei treatedat different drugs concentrations.
3.4. DNAproteins crosslinks
We investigated the ability of Ru compounds toinduce DNAprotein crosslinks by measuring thetotal amount of freely extractable nuclear proteins.Fig. 8 shows the retention of proteins evaluated asdifference of protein concentrations in solution be-tween drug treated and untreated nuclei through the
w xBradford method 14 . The results obtained withCDDP confirm its ability to bind DNA to nuclear
w xproteins 23 . On the other hand, ruthenium com-plexes induce a concentration dependent retention ofproteins very similar to that of CDDP. Unfortu-nately, this experiment does not allow to recognizespecific proteins bound to DNA. Therefore, we canonly suppose that at these relatively high drug con-centrations the binding involves mainly histones.
3.5. Genotoxic actiity on mammalian cells inculture
The cytotoxic activity of the three Ru compoundswas also evaluated on V79 Chinese Hamster cells,by measuring the cloning efficiency of cells plated at
( )A. Barca et al.rMutation Research 423 1999 171181 179
.low density 200 cells per 50 mm dish after treat-ment of confluence culture with the drugs. Fig. 9ashows the surviving fraction of V79 cells plottedagainst drug concentrations: it is noteworthy thatamong the three drugs only TEQU, which on theother hand had shown a very weak interaction withchromatin and DNA, seems to display a relevantcytotoxic activity, with a clear concentration-depen-dent exponential trend. The same compound wasalso tested for the mutagenic activity, showing alsothe ability to induce 6-TG resistant mutants in V79cells, with a linear concentration-dependent trend .Fig. 9b . It should be noted that CDDP has alsobeen studied for its cytotoxic and mutagenic activityon V79 cells showing that this antitumour drug is
y5 w xvery active at concentrations as low as 10 M 24while TEQU, which is the most cytotoxic of thethree compounds tested in this study, reduces thesurviving fraction of less than 10% only at concen-tration as high as 10y3 M. On the other hand, NAMI
.Fig. 9. Cytotoxic and mutagenic activity of drugs on V79 cells. aLog of the surviving fraction; sNAMI, v sNAOX, Bs
. 6TEQU. b Number of mutantsr10 CFC in TEQU-treated cells.Points are the mean values from at least three independent experi-ments.
and NAOX in a range between 1 and 5=10y3 Mare almost devoid of cytotoxicity on V79 cells.
In this paper, we compare three ruthenium com-pounds, which have shown promising antitumourand antimetastatic activity on experimental tumoursw x9 , with the well-known antitumour drug CDDP.Since previous studies had shown that the main
w xtarget for this drug is DNA 1 , we have focused ourattention on the characterization of DNA adductsinduced by the compounds. It is widely accepted thatthe biological activity of antitumour drugs like CDDPcorrelates with their ability to interact with DNA, butthere is a large debate on which adducts are respon-
w xsible for the different biological effects 8 . Manyw xauthors 25,26 suppose that the more relevant effect
of CDDP is the formation of interstrand crosslinks,even if they represent no more than 5% of totalDNADNA crosslinks introduced by the drug;moreover a recent paper on polypyridyl rutheniumcomplexes shows a correlation between cytotoxicityin mammalian cells in culture and interstrand
w xcrosslinks formation 27 .NAMI, NAOX and TEQU bind covalently to
DNA as demonstrated by the inhibition of nucleases,but of them only TEQU induces interstrandcrosslinks. In fact, the renaturation profiles of ther-mally denaturated DNA after treatment with thedrugs show no variation with respect to untreatedDNA, while the experiments with linear pBR322show that in TEQU treated DNA there are inter-strand crosslinks. The discrepancy may be ascribedto the different sensibility between the two methodsinvolved, as demonstrated in the case of the drug
w xangelicin 28,29 , so suggesting that the total numberof interstrand crosslinks is low. In a previous work
w xon the same Ru compounds 16 , the authors hadshown the formation of DNADNA interstrandcrosslinks by all three complexes, in amounts com-parable to that of CDDP. Their experiments were
w xperformed by the Ethidium Bromide method 30 , sowe suggest that the reduced ability of treated DNAto uptake the dye was determined by the presence ofintra-strand crosslinks, which prevent helix extensionand unwinding to accommodate the intercalating
( )A. Barca et al.rMutation Research 423 1999 171181180
molecules. In this view their results are in agreementwith our data on digestion with DNase I, since inboth experiments NAMI and NAOX give resultsvery similar to those of CDDP, while TEQU shows aless pronounced inhibition of the enzyme activity.
Furthermore, CDDP has been previously demon-strated to induce covalent proteinDNA crosslinks:this ability is also confirmed by our data on proteinretention, which demonstrate that also NAMI, NAOXand TEQU are able to crosslink proteins to DNA intreated nuclei. The reaction is supposed to involvemainly histones, even if the retention of specificproteins present in very small amount cannot bedemonstrated by the method used.
It is difficult to find a correlation between theseresults and the demonstrated activity of these drugsagainst tumour cells in vitro and tumour growth and
w xmetastases in vivo 9 . In fact, TEQU displayed asignificant cytotoxic activity against TXL5 lym-phoma cells in vitro but only a weak activity in vivoagainst tumour growth and lung metastases. On thecontrary, NAMI and NAOX are completely devoidof in vitro cytotoxic activity; conversely, they areable to reduce primary tumours growth in vivo andto inhibit even more significantly the formation of
w xlung metastases in mice 9,31 .It should be noted that the cytotoxic activity of
the three Ru compounds against tumour cell lines inw xvitro 9 is consistent with our reported in vitro
cytotoxic activity on V79 cells. Among the threedrugs tested, only TEQU has been shown to becytotoxic on these cells; on the same cell line itshowed also the ability to induce point mutations.We want to stress the fact that TEQU cytotoxicactivity correlates with its ability to induce covalentbridges between the two DNA strands, so reinforcingthe hypothesis that binds cytotoxicity to interstrandcrosslinks formation.
Our results do not approach the effect of reduc-tion of Ru ions. At present they are inadequate tofully explain the biological activity of NAMI, NAOXand TEQU. They indicate DNA as the most relevantbut not the only biological target for the three ruthe-nium compounds and they correlate roughly to theeffects on primary tumours in mice, but they do notprovide any explanation for the antimetastatic activ-ity. In this respect we want to mention that V79cells, after treatment with these drugs for cytotoxic-
ity and mutagenicity assays, became hardly detach- .able from dishes by trypsin data not shown . Of
course, this finding needs further investigation to beexplained, but it could suggest that cell matrixmolecules are modified by the drugs, changing celladhesion properties, which are critical for the metas-tasis of tumour cells; in any case our observationsupports the hypothesis that in addition to DNAother relevant molecules in the cell are involved inthe antitumour activity of ruthenium complexes.
This work was supported by the Italian MinisterodellUniversita e della Ricerca Scientifica.`
w x1 E. Holler, Mechanism of action of tumour-inhibiting metal .complexes, in: B.K. Keppler Ed. , Metal Complexes in
Cancer Chemotherapy, VCH, Weinheim, 1993, pp. 3769.w x2 N. Farrel, Transition Metal Complexes as Drugs and
Chemotherapeutic Agents, Kluwer Academic Publisher, Dor-drecht, 1989.
w x3 G. Mestroni, E. Alessio, G. Sava, S. Pacor, M. Coluccia, Thedevelopment of tumour-inhibiting ruthenium dimethylsulf-
.oxide complexes, in: B.K. Keppler Ed. , Metal Complexesin Cancer Chemotherapy, VCH, Weinheim, 1993, pp. 157185.
w x4 M.J. Clarke, S. Bitler, D. Rennert, M. Buchbinder, A.D.Kelman, Reduction and subsequent binding of rutheniumions catalyzed by subcellular components, J. Inorg. Biochem.
.12 1980 7987.w x5 M.J. Clarke, Progress in Clinical Biochemistry and Medicine,
Vol. 10, Springer-Verlag, Berlin, 1989.w x6 P.D. Lawley, D.H. Phillips, DNA adducts from chemothera-
.peutic agents, Mutat. Res. 355 1996 1340.w x7 M.J. Clarke, Ruthenium complexes: potential roles in anti-
.cancer pharmaceuticals, in: B.K. Keppler Ed. , Metal Com-plexes in Cancer Chemotherapy, VCH, Weinheim, 1993, pp.129156.
w x8 S.E. Sherman, S.J. Lippard, Structural aspects of platinum .anticancer drug interactions with DNA, Chem. Rev. 87 1987
11531181.w x9 G. Sava, S. Pacor, A. Bergamo, M. Cocchietto, G. Mestroni,
E. Alessio, Effects of ruthenium complexes on experimentaltumors: irrelevance of cytotoxicity for metastasis inhibition,
.Chem. Biol. Interact. 95 1995 109126.w x10 E. Alessio, G. Balducci, A. Lutman, G. Mestroni, M. Calli-
garis, W.M. Attia, Synthesis and characterization of two new .classes of ruthenium III sulfoxide complexes with nitrogen
( )A. Barca et al.rMutation Research 423 1999 171181 181
. w . .xdonor ligand L : Na trans-RuCl R SO L and4 2 . . .mer,cisRuCl R SO R SO L : the crystal structure of3 2 2
w . x wNa trans-RuCl DM SO NH P 2DM SO, Na trans-4 3 . x .RuCl DMSO Im P H O P Me CO Im s imidazole and4 2 2
. .mer,cis-RuCl DMSO DMSO NH , Inorg. Chim. Acta 2033 3 .1993 205217.
w x11 J. Sembrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: ALaboratory Manual, CSH Press, New York, 1989.
w x12 J.A. Hartley, R.L. Souhami, M.D. Berardini, Electrophoreticand chromatographic separation methods used to reveal inter-strand crosslinking of nucleic acids, J. Chromatogr. 618 .1993 277288.
w x13 B. Pani, P. Plossi, E. Russo, HMG proteins released from thechromatin following incubation of mammalian nuclei with
.ethidium bromide, Exp. Cell Res. 180 1989 557562.w x14 M.M. Bradford, A rapid and sensitive method for the quanti-
tation of microgram quantities of protein utilizing the princi- .ple of protein-dye binding, Anal. Biochem. 72 1976 248
254.w x15 J.P. ONeill, A.W. Hsie, Phenotypic expression time of
mutagen-induced 6-thioguanine resistance in Chinese ham- .ster ovary cells CHOrHGPRT system , Mutation Res. 59
.1979 109118.w x16 G. Mestroni, E. Alessio, G. Sava, S. Pacor, M. Coluccia, A.
.Boccarelli, Water-soluble ruthenium III dimethyl sulfoxidecomplexes: chemical behaviour and pharmaceutical proper-
.ties, Metal Based Drugs 1 1993 4163.w x .17 J. Hayes, W.M. Scovell, cis-Diamminedichloroplatinum II
modified chromatin and nucleosomal core particle, Biochim. .Biophys. Acta 1089 1991 377385.
w x18 S.E. Sherman, D. Gibson, A.H.-J. Wang, S.J. Lippard, X-raystructure of the major adduct of the anticancer drug cisplatin
w . .x .with DNA: cis- Pt NH d pGpG , Science 230 19853 2412417.
w x19 H.R. Drew, Structural specificities of five commonly used .DNA nucleases, J. Mol. Biol. 176 1984 535557.
w x .20 J.J. Hayes, W.M. Scovell, cis-Diamminedichloroplatinum IImodified chromatin and nucleosomal core particle probed
.with DNase I, Biochim. Biophys. Acta 1088 1991 413418.w x21 K.W. Kohn, C.L. Spears, P. Doty, Inter-strand crosslinking
.of DNA by nitrogen mustard acids, J. Mol. Biol. 19 1966266288.
w x22 M.F. Pera Jr., C.J. Rawlings, J. Shackleton, J.J. Roberts,Quantitative aspects of the formation and loss of DNAinterstrand crosslinks in Chinese hamster cells following
. .treatment with cis-diamminedichloroplatinum II cisplatin :II. Comparison of results from alkaline elution, DNA renatu-ration and DNA sedimentation studies, Biochim. Biophys.
.Acta 655 1981 152166.w x23 S.J. Lippard, J.D. Hoeschele, Binding of cis- and trans-di-
.chlorodiammineplatinum II to the nucleosome core, Proc. .Natl. Acad. Sci. USA 76 1979 60916095.
w x24 R. Chibber, M.J. Ord, The mutagenic and carcinogenic prop-erties of three second generation antitumour platinum com-pounds: a comparison with cisplatin, Eur. J. Cancer Clin.
.Oncol. 25 1989 2733.w x25 V. Brabec, M. Leng, DNA interstrand cross-links of trans-
.diamminedichloroplatinum II are preferentially formed be-tween guanine and complementary cytosine residues, Proc.
.Natl. Acad. Sci. USA 90 1993 53455349.w x26 F. Larminat, W. Zhen, V.A. Bohr, Gene-specific DNA repair
of interstrand cross-links induced by chemotherapeutic agents .can be preferential, J. Biol. Chem. 268 1993 26492654.
w x27 O. Novakova, J. Kasparkova, O. Vrana, P.M. van Vliet, J. Reedijk, V. Brabec, Correlation between cytotoxicity andDNA binding of polypyridyl ruthenium complexes, Biochem-
.istry 34 1995 1236912378.w x28 F. Bordin, F. Carlassare, L. Busulini, F. Baccichetti, Furo-
coumarin sensitization induces DNAprotein cross-links, .Photochem. Photobiol. 58 1993 133136.
w x29 X. Chen, J. Kagan, G. Miolo, F. DallAcqua, D. Averbeck,E. Bisagni, Photosensitized cross-linking and cleavage ofpBR322 and M13 DNA, comparison of 4,4X,6-trimethyl-angelicin and 3-carbethoxypsoralen, J. Photochem. Photo-
.biol. B 22 1994 5157.w x30 A.R. Morgan, D.F. Pulleyblank, Native and denaturated
DNA, cross-linked and palindromic DNA and circular cova-lently-closed DNA analysed by a sensitive fluorometric pro-
.cedure, Biochem. Biophys. Res. Commun. 61 1974 396403.
w x31 G. Sava, S. Pacor, G. Mestroni, E. Alessio, Effects of the . w . x0 wRu III complexes mer-RuCl DMSO Im and Na trans-3 2
. xRuCl DMSO Im on solid mouse tumours, Anti-Cancer4 .Drugs 3 1992 2531.