- Email: [email protected]
Antiproliferative activity of Pt(II) and Pd(II) phosphine complexes with thymine and thymidine
JOURNAL OF
Inorganic Biochemistry Journal of Inorganic Biochemistry 101 (2007) 254–260 www.elsevier.com/locate/jinorgbio
Antiproliferative activity of Pt(II) and Pd(II) phosphine complexes with thymine and thymidine Anna Messere a,1, Enrica Fabbri b,1, Monica Borgatti b, Roberto Gambari b, Benedetto Di Blasio a, Carlo Pedone c, Alessandra Romanelli c,* b
a Dipartimento di Scienze Ambientali, Seconda Universita` di Napoli, via Vivaldi 43, 81100 Caserta, Italy ER-Gen Tech, Dipartimento di Biochimica e Biologia Molecolare, Universita` di Ferrara, via L. Borsari, 46, 44100 Ferrara, Italy c Dipartimento delle Scienze Biologiche Universita` di Napoli ‘‘Federico II’’, via Mezzocannone 16, 80134 Napoli, Italy
Received 2 May 2006; received in revised form 21 September 2006; accepted 22 September 2006 Available online 1 October 2006
Abstract Oxidative addition reactions between [M(PPh3)4] (M = Pt and Pd) and N1-methylthymine (t)/3 0 ,5 0 -di-O-acetylthymidine (T) were carried out to give [M(II)(PPh3)2Cl t (or T)] complexes, in which the metal is coordinated to the N3 of the base. All complexes were characterized by spectroscopic analyses (IR, NMR) and Fast Atom Bombardment mass spectrometry (FAB-MS); X-ray data for the thymine complexes and elemental analysis for the thymidine complexes are reported. The antiproliferative activity of the complexes was tested on human chronic myelogenous leukaemia K562 cells. Arrested polymerase-chain reaction analysis was carried on to correlate antiproliferative activity and inhibition of DNA replication. All Pd and Pt complexes exhibit antiproliferative activity, Pd complexes resulting always more active than Pt complexes. Arrested PCR data are strongly in agreement with the effects on cell growth, suggesting that inhibition of the DNA replication by the synthesized compounds is the major basis for their in vitro antiproliferative activity. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Metal; Antiproliferative; Thymidine; K562
1. Introduction Biological activity of transition metal complexes has been widely investigated, with particular attention to Pt(II) complexes [1–7]. Amine complexes like [Pt(NH3)2LCl] where L is an aromatic ligand and those in which chloride ions are substituted by carboxylates show promising antitumor activity. Based on early studies by Rosenberg showing that cis-diamminedichloroplatinum (cis-DDP) induced filamentous growth in bacteria and inhibited cell division, it was postulated that Pt complexes are mainly involved in DNA replication processes [8]. It has been proposed that Pt complexes interact with DNA after the substitution of one or two chlorine ligands by water molecules, affording *
1
Corresponding author. Tel.: +39 0812536679; fax: +39 0812534574. E-mail address: [email protected] (A. Romanelli). These authors contributed equally to this work.
0162-0134/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2006.09.022
electrophilic species [9]. 1,2 and 1,3 intrastrand crosslinks seem to be responsible for the DNA lesions and therefore for the cytotoxic activity of cis and trans complexes respectively [10]. Several complexes, containing a variety of leaving groups, such as chlorine or any ligand displaced under physiological conditions, and/or carrier ligands, such as amines or molecules which remain bound to the metal, in both cis and trans configuration have been synthesized and tested. It has been shown that changing the leaving groups influences the tissue and intra-cellular distribution of the complexes; steric hindrance of the ligands is also related to the antitumoral activity. Bulky ligands in transPt complexes are proposed to enhance their antitumor activity by limiting the access to the metal atom and slowing down the ligands substitution reaction [11]. It has been demonstrated that in addition to Pt(II) complexes, several complexes of the platinum group exhibit cytostatic activity even in cis-Pt resistant cells [12]. Pd(II) complexes with
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
neutral ligands such as amines pyrimidine, pyridine, pyrazole, aryl groups show antiproliferative and/or antitumor activities [13–15]. Interestingly, dithiocarbamate-amine derivatives of Pd(II) characterized by an antitumor activity comparable to that of cisplatin are able to circumvent cross-resistance to cisplatin [16]. The aim of this work was to study and compare the antiproliferative activity of cis and trans complexes, having a [M(PPh3)2LCl] stoichiometry (M = Pd, Pt) (Schemes 1 and 2). In previous papers [17,18] we described the synthesis of [M(PPh3)2LCl] (L = t, T); as prosecution of our work we investigated the stability of our complexes and their antiproliferative activity. All compounds were obtained by oxidative addition of the [M(PPh3)4] complex on N1methylthymine (t) and 3 0 ,5 0 -di-O-acetylthymidine (T). Reactions with the [Pt(PPh3)4] gave cis complexes, while it is reported that pyridine-like ligands drive the reaction to the formation of trans complexes [10]. Reactions with [Pd(PPh3)4] afforded trans complexes. It is reported that metals belonging to the same group, such as Pt and Pd, have similar chemical properties. Studies on Pt and Pd complexes show the differences between Pd and Pt complexes are of kinetic nature [19,20]. Considering the similarity in the chemical properties of platinum and palladium ions, we compared Pt and Pd complexes assuming that the different behaviour would be mainly due to the difference in the cis or trans geometry of the complexes. [M(PPh3)4] precursors were used in the complexation reactions since it is reported that phosphine ligands by themselves and phosphine complexes of other metals, such as Ag(I), Au(I) and Sn(IV) are anticancer, anti-HIV or antimitochondrial agent [21–23]. Furthermore thymine and thymidine ligands were chosen since they are natural
255
components of DNA, and have been shown alter the pharmacological activity compound complexed with them. 2. Materials and methods Chemicals were obtained from commercial suppliers (Fluka) and used without further purification. The 1H and 13C NMR spectra were recorded on a Bruker WM400 spectrometer at 400 and 100.13 MHz respectively. All chemical shifts are expressed in ppm with respect to the signal of the protonated solvent (CD3OD: d 3.31 and 49.5; DMSO-d6: d 2.55 and 39.5). The 31P NMR spectra were run on a Bruker WM-400 spectrometer at 161.98 MHz, with external reference to 85% H3PO4 (d 0.0). The Fast Atom Bombardment (FAB) mass spectra (positive) were recorded on a ZAB 2SE spectrometer. The IR spectra were recorded on a Perkin–Elmer 457 spectrophotometer. Elemental analyses were performed by the ‘‘Analytische Laboratorien Prof. Dr. H. Malissa und G. Reuter GmbH’’ Industriepark Kaiserau (Haus Heidbruch) D-51789 Lindlar Germany. 2.1. Crystallography Intensity data collection was performed using graphite˚ ) and a monochromated Mo-Ka radiation (k = 0.70930 A pulse-high discrimination on an automated MACH3 Enraf-Nonius for complex 1. Intensity data collection for complex 2 was performed using graphite-monochromated ˚ ) and a pulse-high discrimCu-Ka radiation (k = 1.54178 A ination on a CAD4 Enraf-Nonius. Data analysis was performed as described in previous papers [17,18]. 2.2. Synthesis of complexes 1 and 2
Scheme 1.
Scheme 2.
1 equivalent of 1-methylthymine was reacted with 1 equivalent of [M(PPh3)4] (M = Pt, Pd) in refluxing dry toluene in the presence of KCl (1 equivalent). After stirring 5 h, toluene was evaporated. The crude mixture was redissolved in 7:3 v/v C6H6/CH3COOCH2CH3 and purified by silica gel chromatography. 1: Pd complex containing 1-methylthymine: Yield: 59%; Rf 0.4 (eluent C6H6/CH3COOCH2CH3 7/3 v/v). 1H NMR: (DMSO-d6) d 7.8–7.4 (15H, C6H5–P, m); 6.3 (1H, H6, bs); 2.6 (3H, CH3–N1, s); 1.2 (3H, CH3–C5, bs). 13C NMR: (CDCl3) d 170.1 (C4); 154.7 (C2); 134.8–127.7 (C6H5–P); 107.9 (C5); 36.1 (CH3–N1); 13.1 (CH3–C5). 31P NMR: (DMSO-d6) d 23.5. FAB-MS: (106Pd) 771m/z; [M + H]+. IR [(CHCl3) (m cm1)]: 1670, 1640 (C@O); 510 (PdAN). Molecular structure of trans-[PdCl(1-methylthymine) (PPh3)2] Æ H2O, 1 [18]. The palladium atom displays square planar coordination: two trans corners of the square plane are occupied by the phosphorus atoms of two triphenylphosphines. The chlorine atom and the amide nitrogen of the 1-methylthymine ligand are trans. Crystal data. C42H37ClN2O2P2Pd Æ H2O, MW = 822.01, monoclinic, space group name P21/c (no. 14),
256
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
˚ , b = 12.955(6) A ˚ , c = 19.190(10) A ˚, b= a = 17.780(10) A 3 ˚ 117.08(5)°, V = 3936(4) A , T = 298 K, Z = 4, l(Mo Ka) = 0.794 mm1, 7473 unique reflections (Rint = 0.0) used in all calculations. The final Rw(F2) was 0.1908; R1 = 0.0707. 2: Pt complex containing 1-methylthymine: Yield: 82%. Rf 0.44 (eluent CHCl3–MeOH 97:3, v/v). 1 H NMR (DMSO-d6): d 8.20–7.00 (30H, phenyl protons, m); 6.31 (1H, H-6, s); 3.0 (3H, CH3-1, s); and 1.62 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6): d 169.9 (C4); 155.1 (C-2); 138.6 (C-6); 109.3 (C-5); 36.1 (CH3-1); and 12.1 (CH3-5). 31P NMR (DMSO-d6): d 15.4 [d, PPh3 trans to Cl, 1J(Pt–P) = 3984] and 8.6 [d, PPh3 cis to Cl, 1 J(Pt–P) = 3235 Hz]. FAB MS (195Pt, 35Cl): m/z 893, [M + H]+ and 957 [M + Cl]+. IR [(CHCl3) (m cm1)]: 1690, 1640 (C@O). Molecular structure of cis-[PtCl(1-MeThy)(PPh3)2] Æ MeOH, 2 [17]. The platinum atom displays square planar coordination: two cis corners of the square plane are occupied by the P atoms of two triphenylphosphines and the chlorine atom and the amide nitrogen of the methylthymine ligand are in cis position to each other. The four atoms co-ordinated to the metal lie in a plane with very ˚ ) deviation from it. The plane of small (less than 0.08 A the methylthymine ligand is approximately perpendicular to the platinum coordination plane. Crystal data. C42H37ClN2O2P2Pt Æ CH3OH, M = 923.23, ˚, b= triclinic, space group P 1 (no. 2), a = 11.176(8) A ˚ , c = 13.36(1) A ˚ , a = 97.31(6)°, b = 91.32(6)°, 13.892(9) A c = 88.00(6)°, U = 2056(3) A3, T = 293 K, Z = 2,l(Cu-Ka) = 8.029 mm1, 7789 unique reflections (Rint = 0.0) used in all calculations. The final wR(F2) was 0.1334; R1 = 0.0515. 2.3. Synthesis of complexes 3, 4a and 4b Procedure described for 1 using as substrate. 3 0 ,5 0 -di-Oacetylthymidine. 3: Pd complex containing 3 0 ,5 0 -di-O-acetylthymidine: Yield: 87%. Rf 0.7 (eluent CHCl3/CH3OH 98:2 v/v). 1H NMR: (DMSO-d6) d 7.7–7.5 (15H, C6H5–P, m); 6.4 (1H, H6, bs); 5.8 (1H, H1 0 , dd); 5.1 (1H, H3 0 , m); 4.2 (2H, H5 0 , m); 4.1 (1H, H4 0 , m); 2.5 (2H, H2 0 , m); 2.1 (3H, CH3–CO, s); 2.0 (3H, CH3–CO, s); 1.2 (3H, CH3–C5, bs). 13C NMR: (CDCl3) d 169.8 and 170.1 (CH3–CO); 153.7 (C2); 168.9 (C4); 134.7–127.71 (C6H5–P); 84.2 (C1 0 ); 108.9 (C5); 80.7 (C3 0 ); 73.8 (C4 0 ); 63.4 (C5 0 ); 36.7 (C2 0 ); 20.4, 20.6 (CH3–CO); 13.4 (CH3–C5). 31P NMR: (DMSO-d6) d 23.5. FAB-MS: (106Pd) 991 m/z; [M + H]+. IR [(CHCl3) (m cm1)]: 1769, 1690, 1660 (C@O); 520 (PdAN). Anal. Calcd for C50H47ClN2O7P2Pd: C, 60.55; H, 4.78; Cl, 3.57; N, 2.82; P, 6.25; Found: C, 59.69; H, 4.60; Cl, 3.92; N, 3.10; P,7.03. 4a and 4b: Pt complexes containing 3 0 ,5 0 -di-O-acetylthymidine : Yield of the mixture: 80%. Rf 0.4 (eluent CHCl3–MeOH 97:3, v/v).
Complex 4a: 1H NMR (DMSO-d6) d 8.32–7.00 (30H, phenyl protons, m), 6.67 (1H, H-6, s), 6.17 (1H, H-1 0 , dd), 5.14 (1H, H-3 0 , m), 4.23 (2H, H-5 0 , m), 4.10 (1H, H4 0 , m), 2.14 and 2.07 (3H each, CH3CO, s), 2.1–1.9 (2H, H-2 0 , m) and 1.47 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6): d 170.0 (CH3CO), 167.9 (C-4), 153.6 (C-2), 134.7–127.5 (phenyl carbons and C-6), 110.1 (C-5), 83.6 (C-49), 80.4 (C-19), 73.2 (C-39), 63.6 (C-59), 35.4 (C-29), 20.7 and 20.5 (2CH3CO) and 13.3 (CH3-5). 31P NMR (DMSO-d6) d 15.3 [d, PPh3 trans to Cl, 1J(Pt–P) = 3960] and 8.6 [d, PPh3 cis to Cl, 1J(Pt–P) = 3265 Hz]. Complex 4b: 1H NMR (DMSO-d6) d 8.30–7.00 (30H, phenyl protons, m), 6.64 (1H, H-6, s), 6.26 (1H, H-1 0 , dd), 5.19 (1H, H-3 0 , m), 4.26 (2H, H-5 0 , m), 4.13 (1H, H-4 0 , m), 2.25 (2H, H-2 0 , m), 2.14 and 2.11 (3H each, CH3CO, s) and 1.48 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6) d 170.0 (CH3CO), 167.8 (C-4), 154.1 (C-2), 134.7–127.5 (phenyl carbons and C-6), 109.2 (C-5), 83.6 (C-4 0 ), 80.3 (C-1 0 ), 74.2 (C-3 0 ), 63.6 (C-5 0 ), 35.3 (C-2 0 ), 20.7 and 20.5 (2 CH3CO) and 13.3 (CH3-5). 31P NMR (DMSO-d6) d 15.3 [d, PPh3 trans to Cl, 1J(Pt–P) = 3960] and 8.6 [d, PPh3 cis to Cl, 1J(Pt–P) = 3265 Hz]. FAB MS on mixture (195Pt, 35 Cl): m/z 1044, [M–Cl]+ and 754 [M-nucleoside]+. IR [(CHCl3) (m cm1)] on mixture 1750, 1663, 1590 [strong, m(CO)], 308 cm1 [weak, m(Pt–Cl)]. Anal. Calcd for C50H47ClN2O7P2Pt: C, 55.58; H, 4.38; Cl, 3.28; N, 2.59; P, 5.73; Found: C, 55.70; H, 4.42; Cl, 3.15; N, 2.30; P, 5.65. 2.4. Synthesis of complexes 5 and 6 0.1 g of derivative 3 (or 4) were dissolved in 2 mL CH3OH; 2 mL of NH4OH (37%) were added. The mixture was treated at 50 °C 2 h; the solvent was evaporated and the product was purified by silica gel column chromatography. 5: Pd complex containing thymidine: Yield:60%. Rf 0.8 (eluent 85:15 CHCl3/CH3OH v/v). 1 H NMR: (CD3OD) d 7.9–7.4 (15H, C6H5–P, m); 6.9 (1H, H6, bs); 5.7 (1H, H1 0 , dd); 4.2 (1H, H3 0 , m); 3.8 (1H, H4 0 , m); 3.7 (2H, H5 0 , m); 2.1 (1H, H2 0 , m); 1.7 (1H, H2 0 , m); 1.4 (3H, CH3-C5, bs). 13C NMR: (CD3OD) d 172.3 (C4); 156.1 (C2); 136.2–129.6 (C6H5–P); 110.2 (C5); 87.1 (C1 0 ); 88.5 (C4 0 ); 71.9 (C3 0 ); 63.1 (C5 0 ); 41.7 (C2 0 ); 13.9 (CH3-C5). 31P NMR (CD3OD): d 26.6 IR [CHCl3 (m cm1)]: 1667, 1646 (C@O); 520 (Pd–N). 6a and 6b mixture: Pt complexes containing thymidine: Yield: 90%. Rf 0.2 (eluent 85:15 CHCl3/CH3OH v/v). 1H NMR (CD3OD): d 7.80–7.10 (60H, phenyl protons and H-6, m), 6.19 and 6.01 (1H each, H-1 0 , dd), 4.33 (2H, H3 0 , m), 3.87 (2H, H-4 0 , m), 3.72 (4H, H-5 0 , m) 2.30–2.00 (4H, H-2 0 , m), 1.69 and 1.62 (3H each, CH3-5, s). 13C{1H} NMR (CD3OD): d 172.7 (C-4), 156.9 (C-2), 137– 128.0 (phenyl carbons and C-6), 111.9 and 108.7 (C-5), 88.8 (C-4 0 ), 87.5 and 86.9 (C-1 0 ), 72.5 and 72.4 (C-3 0 ), 63.2 (C-5 0 ), 41.8 and 41.4 (C-2 0 ) and 13.8 (CH3-5). 31P NMR (CD3OD): d 13.5 [d, PPh3 trans to Cl, 1J(Pt– P) = 3971] and 6.8 [d, PPh3 cis to Cl, 1J(Pt–P) = 3305
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
Hz]; IR [(CHCl3) (m cm1)] 1655, 1578 [strong, m(CO)], 300 cm1 [weak, m(Pt–Cl)]. 3. Results
257
Pt complexes 6a and 6b. Pt complexes 4a and 4b resulted more stable than Pd complex 5. Free thymidine nucleoside was isolated after deprotection of 5 by NH3. No cis–trans isomerization was observed in basic ammonia medium, as demonstrated by 31P NMR spectra.
3.1. Chemistry 3.2. Biological assays Complexes 1–4 were synthesized reacting [M(PPh3)4] (M = Pt and Pd) complexes and KCl respectively with N1-methylthymine and 3 0 ,5 0 -di-O-acetylthymidine in refluxing toluene, following procedures reported in the literature [17,18]. All reactions afforded M(II) metal complexes, characterized by an M-N3 of thymine or thymidine bond. Structure of the complexes was determined by NMR, IR, FAB-MS; X-ray data for complexes 1 and 2 and elemental analysis for complexes 3, 4a and 4b are reported. Reaction of [Pt(PPh3)4] with 3 0 ,5 0 -di-O-acetylthymidine afforded two diastereoisomeric species (4a and 4b), rapidly interconverting one into the other, whose presence can be ascribed both to the restricted rotation around the Pt–N3 bond and to the presence of fixed configurations at the sugar carbons [17]. When the same reaction was carried out with [Pd(PPh3)4] no diastereoisomers were isolated, because in this case the rotation around the Pd–N3 is very fast [18]. 1H and 31P NMR spectra of 3 recorded in CDCl3 at low temperature (15 °C) showed always just one signals pattern confirming no diastereoisomers can be isolated. All Pt complexes are characterized by the two triphenylphosphine in a cis geometry as demonstrated by the presence of two signals in the 31 P NMR, while Pd analogues have trans triphenylphosphines (just one signal in the 31P NMR spectrum) and the chloride ion trans to the base N3 (Schemes 1 and 2). Xray analysis performed on thymine complexes 1 and 2, revealed the presence of distorted square planar complexes. To further characterize Pd and Pt complexes, we performed ligand exchange experiments by NMR. Ligands are not displaced by an S donor such as DMSO as demonstrated by 1H and 31P NMR experiments in DMSO on complexes 3, 4a and 4b. In fact we did not observe any change in the ligand signals in the NMR spectra recorded at different times and scanning the temperature up to 50 °C. Furthermore treatment of 3 with a very reactive sulphorated molecule, such as CS2, did not cause ligand exchange, even after few days reaction [18]. More exchange experiments were carried out in excess of D2O (O donor) and of thymidine (N, O donor). We did not observe in the 1H and 31P NMR spectra neither a shift nor a variation in the integration for the signals of the ligands, which could have been due to the presence of displaced molecules. With the aim to obtain complexes soluble in aqueous media we deblocked the hydroxyl functions on deoxyribose moiety by ammonia treatment. This experiment was useful for testing the stability of the complexes to NH3 exchange. Treatment of complexes 3, 4a and 4b with NH3(aq) 37% at 50 °C gave complexes 5, 6a and 6b, in which the 3 0 and 5 0 position of the sugar have been deacetylated with yields ranging from 40% for the Pd complex 5 to 90% for the
Arrested PCR was performed as an assay for determining the biological activity in vitro [24,25]. This assay has been demonstrated useful for preliminary screening of DNA-binding drugs [24–26]. Briefly, target DNA was pre-incubated for 15 min with increasing concentrations of cisplatin or analyzed compounds and then PCR reaction was conducted with the 5 0 -TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-3 0 (forward) and 5 0 -CTA GAA GCA TTT GCG GTG GAC GAT GGA GGG-3 0 (reverse) primers, leading to an amplification of a 550 bp PCR product of b-actin, which was analyzed by agarose gel electrophoresis. PCR conditions were (20 cycles): denaturation 94 °C, 1 min; annealing 63 °C, 1 min; elongation: 72 °C, 1 min. The antiproliferative effect was analyzed on the human chronic myelogenous K562 cell line [27], obtained from American Type Culture Collection (ATCC, Rockville, MD) and grown in vitro in RPMI 1640 medium (Sigma/ Aldrich, Milwaukee, WI, USA), supplemented with 10% fetal bovine serum (FBS) (Celbio, Milano, Italy), 2 mM L-glutamine (Sigma/Aldrich, Milwaukee, WI, USA), a solution of 50 U/mL penicillin and 50 lg/mL streptomycin (Sigma/Aldrich, Milwaukee, WI, USA) in 5% CO2/air humidified atmosphere at 37 °C [28,29]. K562 cells were usually seeded at 30.000/mL and their treatment with compounds 1–6, triphenylphosphine (PPh3) and tryphenylphosphine oxide (PPh3O) was carried out by adding the appropriate drug concentrations at the beginning of the cultures. All the compounds were dissolved in MeOH. Control cultures containing the vehicle were always run in parallel to demonstrate absence of biological effects of MeOH under the experimental conditions employed. The medium was not changed during the induction period. To determine the antiproliferative activity of these compounds, exponentially growing cells were continuously exposed to various concentrations of drugs for 72 h and cell growth was studied by determining the cell number/ mL using Z1 Coulter Counter (Coulter Electronics, Hialeah, FL, USA) [29]. A representative example of arrested PCR data obtained is shown in Fig. 1 and the results from three different experiments summarized in Table 1, which suggests that important differences exist when the activities of the tested compounds are compared each other and with that of cisplatin. The most active Pd-compounds is compound 5, the most active Pt-compound is compound 6. When structurally related Pd- and Pt-compounds are compared (compare compound 3 with 4 and compound 5 with 6), it appears that Pd-compounds are more active than Ptcompounds in inhibiting PCR amplification.
258
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
Fig. 1. Effects of compounds 1–6, cisplatin, PPh3 and PPh3O on generation of PCR products. Target genomic DNA was pre-incubated for 15 min with increasing lM concentrations of compounds, as indicated and then PCR reaction was conducted with the 5 0 -TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-3 0 (forward) and 5 0 -CTA GAA GCA TTT GCG GTG GAC GAT GGA GGG-3 0 (reverse) primers. PCR products were analyzed by agarose gel electrophoresis and ethidium-bromide staining (M: marker). Table 1 Effects of compounds 1–6, triphenylphosphine and triphenylphosphine oxide on polymerase chain reaction (PCR) and cell growth of K562 cells Compound
Inhibition of PCR (IC50, lM)
Inhibition of cell growth (IC50, lM)
Cisplatin 1 2 3 4 5 6 Triphenylphosphine Triphenylphosphine oxide
8.5 ± 1.5 15.0 ± 5.8 25.5 ± 3.5 3.5 ± 0.8 7.0 ± 1.7 1.5 ± 0.3 3.1 ± 0.9 >50 >50
2.10 ± 0.91 4.25 ± 1.10 28.25 ± 8.20 2.25 ± 0.62 3.73 ± 0.81 0.58 ± 0.12 2.95 ± 1.11 5.07 ± 1.00 >100
A representative experiment on the effects on K562 cell growth is reported in Fig. 2 and the results of four independent experiments are shown in Table 1 and expressed as IC50 (dose causing 50% inhibition of cell growth in treated cultures relative to untreated controls). For each drug concentration, duplicate cultures were used. Vehicle or solvent controls were run in each experiment. The results on inhibition of K562 cell growth demonstrate that only one molecule is inactive (PPh3O). As found in arrested PCR experiments, Pd-compounds (among which the most active is compound 5) are more active than Pt-compounds. 4. Discussion The antiproliferative activity of cis and trans[M(PPh3)2LCl] complexes was tested on human
Fig. 2. Effects of compounds 1–6, PPh3 and PPh3O on K562 cell growth. K562 cells were cultured for 3 days in the presence of the indicated concentrations of compounds and then the cell number/ml was determined. d, compound 1; s, compound 2; j, compound 3; h, compound 4; m, compound 5; n, compound 6; r, PPh3; and e, PPh3O.
chronic myelogenous K562 cells. Based on early studies on bifunctional Pt cis and trans complexes it was postulated that the biological activity of metal complexes is due to their interaction with DNA; drug interstrand and intrastrand crosslinking cause lesion to DNA which inhibit its replication. It is now clear that cis and trans complexes interact with DNA in a different way; interstrand and intrastrand linkages are mainly mediated by guanine pairs. Square planar d8 Pt complexes have been shown to intercalate into DNA trough p–p and electrostatic interactions
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
[30], while Hoechst derivatives bind to the minor groove [31]. Monofunctional adduct, either cis and trans, are also able to bind DNA [32,33] and to terminate in vitro DNA and RNA synthesis [34]. Sterically hindered Pt complexes, such as [Pt(amine)2hetCl]+ (het: N-heterocyclic ligand) bind to DNA after one amine ligand, trans to the chloride has been lost [35]. In this work, planar aromatic ligands such as thymine or thymidine were used as metal ligands for their ability to interact with nucleic acids by stacking between two adjacent bases of a DNA strand. The presence of thymine should help the binding between the metal complex and the nucleic acids, introducing an additional binding site and directing the complex also toward adenine containing sequences. The steric hindrance of the ligands should slow down the reaction rate with the DNA and probably affect the reaction mechanism and the selectivity for the binding sites [35]. In addition it is reported that hindered complexes are used for circumventing resistance by sterically inhibiting cellular detoxification by glutathione and other cellular thiols [36]. The presence of at least one leaving group in the synthesized complexes should allow the reaction with DNA, either for the cis and the trans complexes, as demonstrated in the literature [35]. In previous papers we investigated the reactivity of [M(PPh3)] (M = Pd, Pt) complexes toward DNA bases, such as thymidine and thymine. Now we are interested in the characterization of the antiproliferative activity of the complexes containing t/T and Pd/Pt in cells. In order to investigate the stability of ligands in the described complexes we started studying the reactivity of our complexes in ligand exchange reactions. It is in fact known that ligand exchange reactions occurring in cells by thiols such as glutathione, inactivate metal complexes. S-donors ligands, such as DMSO, CS2, O-donors such as H2O, N,O-donors such as thymine, were used in exchange experiments. Reactions were followed by 1H and 31P NMR; the position and integration of the signals relative to bound ligands was not modified by ligands added, suggesting no exchange occurred. Based on these results we proceeded to the study of the antiproliferative activity of complexes in K562 cells. Analysis of the antiproliferative activity of Pt and Pd complexes reveals it is related to the hydrophilicity of the complexes: N1-methylthymine complexes 1 and 2 are less active than thymidine complexes 5, 6a and 6b. Complexed acetylated nucleosides (3 and 4) show the same IC50 no matter what the metal is, while for the corresponding free O-acetyl complexes 5 and 6 IC50 are significatively different, Pd complex resulting more active than Pt complex (Table 1). Reasonably these results can be explained considering the greater up-take of trans complexes; it is in fact reported that trans-Pt complexes are better internalized into cells than cis complexes [37]. Likely, DNA adducts formed by cis and trans compounds are different. The difference in the reaction rate of Pd and Pt complexes (Pd complexes are known for their lability in ligand exchange [15]), in this case should not affect the DNA com-
259
plexation rate. In order to determine whether the biological activity of the synthesized compounds was due to the entire complexes and not to lost ligands, the activity of phosphine ligands and phosphine oxide was tested; it is known that phosphine complexes lose ligands and that in oxidative conditions phosphines go to phosphine oxide. IC50 for phosphines was found 5.07 ± 0.95, while for phosphine oxide it was >100. Although in experimental NMR conditions the complexes resulted stable, it is not possible to exclude that exchange reaction occur in cells with release of phosphines. The IC50 measured for PPh3 results comparable to that of complexes 1, 3, 4 and 6, thus not allowing us to say whether the activity of the complexes can be attributed to the phosphines released or to the complexes as they are. With the exception of PPh3, arrested PCR data are strongly in agreement with the effects on cell growth, suggesting that differential inhibition of DNA replication by the tested compounds fully explains their in vitro antiproliferative activity. Additional mechanism(s) of action should be the molecular basis of the effects of PPh3 on K562 cell growth. 5. Conclusions Monofunctional, sterically hindered, cis and trans complexes [M(PPh3)2LCl] were synthesized by oxidative addition of [M(PPh3)4] to thymine and thymidine. The stability of the complexes was tested and confirmed by NMR ligands exchange experiments. All complexes were tested in antiproliferative assay on the human chronic myelogenous K562 cell line. Analyses of IC50 show that the antiproliferative activity: – is due to DNA binding as suggested by PCR inhibition, – is related to the hydrophilicity of the complexes, which probably affect their biodistribution and biodisponibility, – is strongly correlated with inhibition of PCR reactions and – is higher for the trans free O-acetyl complex 5. Furthermore, in the hypothesis all complexes 1–6 interact with DNA and ligand exchange does not occur as demonstrated by our in vitro experiments, it is likely that DNA lesions repair, which takes place in vivo by substitution of ligands with thiourea and glutathione, would not be observed because of the high stability of the complexes. Acknowledgment RG is funded by AIRC. References [1] D. Wang, S.J. Lippard, Nat. Rev. Drug Discov. 4 (2005) 307–320. [2] B. Rosenberg, L. VanCamp, J.E. Trosko, V.H. Mansour, Nature 222 (1969) 385–386.
260
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
[3] S.J. Lippard, Science 218 (1982) 1075–1082. [4] E. Monti, M. Gariboldi, A. Maiocchi, E. Marengo, C. Cassino, E. Gabano, D. Osella, J. Med. Chem. 48 (2005) 857–866. [5] W. Friebolin, G. Schilling, M. Zoller, E. Amtmann, J. Med. Chem. 47 (2004) 2256–2563. [6] L. Giovagnini, L. Ronconi, D. Aldinucci, D. Lorenzon, S. Sitran, D. Fregona, J. Med. Chem. 48 (2005) 1588–1595. [7] N. Bianchi, F. Ongaro, C. Chiarabelli, L. Gualandi, C. Mischiati, P. Bergamini, R. Gambari, Biochem. Pharmacol. 60 (2000) 31–40. [8] B. Rosenberg, L. Vancamp, T. Krigas, Nature 205 (1965) 698–699. [9] A.L. Pinto, S.J. Lippard, Biochim. Biophys. Acta 780 (1985) 167–180. [10] G. Natile, M. Coluccia, Coord. Chem. Rev. 216–217 (2001) 383–410. [11] N. Farrell, T.T. Ha, J.P. Souchard, F.L. Wimmer, S. Cros, N.P. Johnson, J. Med. Chem. 32 (1989) 2240–2241. [12] M.J. Clarke, F. Zhu, D.R. Frasca, Chem. Rev. 99 (1999) 2511–2534. [13] J. Kuduk-Jaworska, A. Puszko, M. Kubiak, M. Pelczynska, J. Inorg. Biochem. 98 (2004) 1447–1456. [14] G. Zhao, L. Huakuan, P. Yu, H. Sun, S. Zhu, X. Su, Y. Chen, J. Inorg. Biochem. 73 (1999) 145–149. [15] E. Budzisz, U. Krajewska, M. Rozalski, A. Szulawska, M. Czyz, B. Nawrot, Eur. J. Pharm. 502 (2004) 59–65. [16] V. Alverdi, L. Giovagnini, C. Marzano, R. Seraglia, F. Bettio, S. Sitran, R. Graziani, D. Fregona, J. Inorg. Biochem. 98 (2004) 1117– 1128. [17] L. De Napoli, R. Iacovino, A. Messere, D. Montesarchio, G. Piccialli, A. Romanelli, F. Ruffo, M. Saviano, J. Chem. Soc., Dalton Trans. (1999) 1945–1949. [18] A. Romanelli, R. Iacovino, G. Piccialli, F. Ruffo, L. De Napoli, C. Pedone, B. Di Blasio, A. Messere, Organometallics 24 (2005) 3401– 3406. [19] S. Gonzalez, J.M. Tersero, A. Matilla, J.M. Perez, V.M. Gonzalez, C. Alonso, J. Niclos-Gutirriez, J. Inorg. Biochem. 61 (1996) 261–272.
[20] U. Frey, S. Elmroth, B. Modlet, L.I. Elding, A.E. Merbach, Inorg. Chem. 30 (1991) 5033–5037. [21] S.J. Berners-Price, P.J. Dasler, Coord. Chem. Rev. 151 (1996) 1–40. [22] Y. Inouye, T. Kanamori, T. Yoshida, X. Bu, M. Shionoya, T. Koike, E. Rimura, Biol. Pharm. Bull. 17 (1994) 243–250. [23] H. Mitsuya, S. Broder, Nature 325 (1987) 773–778. [24] M. Passadore, N. Bianchi, G. Feriotto, C. Mischiati, P. Giacomini, R. Piva, R. Gambari, Biochem. J. 308 (1995) 513–519. [25] M. Passadore, N. Bianchi, G. Feriotto, C. Mischiati, C. Rutigliano, R. Gambari, Eur. J. Pharmacol. 319 (1997) 317–325. [26] P.G. Baraldi, B. Cacciari, R. Romagnoli, G. Spalluto, R. Gambari, N. Bianchi, M. Passadore, P. Ambrosino, N. Mongelli, P. Cozzi, C. Geroni, Anticancer Drug Des. 12 (1997) 555–576. [27] B.B. Lozzio, C.B. Lozzio, Int. J. Cancer 19 (1977) 136. [28] E. Fibach, N. Bianchi, M. Borgatti, E. Prus, R. Gambari, Blood 102 (2003) 1276–1281. [29] P.G. Baraldi, R. Romagnoli, N. Bianchi, R. Gambari, Bioorg. Med. Chem. 11 (2003) 2381–2388. [30] E.R. Jamieson, S.J. Lippard, Chem. Rev. 99 (1999) 2467246–2467298. [31] M.A. Carrondo, M. Coll, J. Aymami, A.H. Wang, G.A. van der Marel, J.H. van Boom, A. Rich, Biochemistry 28 (1989) 7849–7859. [32] V. Brabec, V. Boudny, Z. Balcarova, Biochemistry 33 (1994) 1316– 1322. [33] M.B. Janik, B. Lippert, J. Bio. Inorg. Chem. 4 (1999) 645–653. [34] Z. Balcarova, J. Kasparkova, A. Zakovska, O. Novakova, M.F. Sivo, G. Natile, V. Brabec, Mol. Pharmacol. 53 (1998) 846–855. [35] M.B. Janik, A. Hegmans, E. Freisinger, B. Lippert, J. Bio. Inorg. Chem. 4 (1999) 130–139. [36] T. Okada, I.M. El-Mehasseb, M. Kodaka, T. Tomohiro, K. Okamoto, H. Okuno, J. Med. Chem. 44 (2001) 4661–4667. [37] N. Farrell, L.R. Kelland, J.D. Roberts, M. Van Beusichem, Cancer Res. 52 (1992) 5065–5072.
Inorganic Biochemistry Journal of Inorganic Biochemistry 101 (2007) 254–260 www.elsevier.com/locate/jinorgbio
Antiproliferative activity of Pt(II) and Pd(II) phosphine complexes with thymine and thymidine Anna Messere a,1, Enrica Fabbri b,1, Monica Borgatti b, Roberto Gambari b, Benedetto Di Blasio a, Carlo Pedone c, Alessandra Romanelli c,* b
a Dipartimento di Scienze Ambientali, Seconda Universita` di Napoli, via Vivaldi 43, 81100 Caserta, Italy ER-Gen Tech, Dipartimento di Biochimica e Biologia Molecolare, Universita` di Ferrara, via L. Borsari, 46, 44100 Ferrara, Italy c Dipartimento delle Scienze Biologiche Universita` di Napoli ‘‘Federico II’’, via Mezzocannone 16, 80134 Napoli, Italy
Received 2 May 2006; received in revised form 21 September 2006; accepted 22 September 2006 Available online 1 October 2006
Abstract Oxidative addition reactions between [M(PPh3)4] (M = Pt and Pd) and N1-methylthymine (t)/3 0 ,5 0 -di-O-acetylthymidine (T) were carried out to give [M(II)(PPh3)2Cl t (or T)] complexes, in which the metal is coordinated to the N3 of the base. All complexes were characterized by spectroscopic analyses (IR, NMR) and Fast Atom Bombardment mass spectrometry (FAB-MS); X-ray data for the thymine complexes and elemental analysis for the thymidine complexes are reported. The antiproliferative activity of the complexes was tested on human chronic myelogenous leukaemia K562 cells. Arrested polymerase-chain reaction analysis was carried on to correlate antiproliferative activity and inhibition of DNA replication. All Pd and Pt complexes exhibit antiproliferative activity, Pd complexes resulting always more active than Pt complexes. Arrested PCR data are strongly in agreement with the effects on cell growth, suggesting that inhibition of the DNA replication by the synthesized compounds is the major basis for their in vitro antiproliferative activity. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Metal; Antiproliferative; Thymidine; K562
1. Introduction Biological activity of transition metal complexes has been widely investigated, with particular attention to Pt(II) complexes [1–7]. Amine complexes like [Pt(NH3)2LCl] where L is an aromatic ligand and those in which chloride ions are substituted by carboxylates show promising antitumor activity. Based on early studies by Rosenberg showing that cis-diamminedichloroplatinum (cis-DDP) induced filamentous growth in bacteria and inhibited cell division, it was postulated that Pt complexes are mainly involved in DNA replication processes [8]. It has been proposed that Pt complexes interact with DNA after the substitution of one or two chlorine ligands by water molecules, affording *
1
Corresponding author. Tel.: +39 0812536679; fax: +39 0812534574. E-mail address: [email protected] (A. Romanelli). These authors contributed equally to this work.
0162-0134/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2006.09.022
electrophilic species [9]. 1,2 and 1,3 intrastrand crosslinks seem to be responsible for the DNA lesions and therefore for the cytotoxic activity of cis and trans complexes respectively [10]. Several complexes, containing a variety of leaving groups, such as chlorine or any ligand displaced under physiological conditions, and/or carrier ligands, such as amines or molecules which remain bound to the metal, in both cis and trans configuration have been synthesized and tested. It has been shown that changing the leaving groups influences the tissue and intra-cellular distribution of the complexes; steric hindrance of the ligands is also related to the antitumoral activity. Bulky ligands in transPt complexes are proposed to enhance their antitumor activity by limiting the access to the metal atom and slowing down the ligands substitution reaction [11]. It has been demonstrated that in addition to Pt(II) complexes, several complexes of the platinum group exhibit cytostatic activity even in cis-Pt resistant cells [12]. Pd(II) complexes with
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
neutral ligands such as amines pyrimidine, pyridine, pyrazole, aryl groups show antiproliferative and/or antitumor activities [13–15]. Interestingly, dithiocarbamate-amine derivatives of Pd(II) characterized by an antitumor activity comparable to that of cisplatin are able to circumvent cross-resistance to cisplatin [16]. The aim of this work was to study and compare the antiproliferative activity of cis and trans complexes, having a [M(PPh3)2LCl] stoichiometry (M = Pd, Pt) (Schemes 1 and 2). In previous papers [17,18] we described the synthesis of [M(PPh3)2LCl] (L = t, T); as prosecution of our work we investigated the stability of our complexes and their antiproliferative activity. All compounds were obtained by oxidative addition of the [M(PPh3)4] complex on N1methylthymine (t) and 3 0 ,5 0 -di-O-acetylthymidine (T). Reactions with the [Pt(PPh3)4] gave cis complexes, while it is reported that pyridine-like ligands drive the reaction to the formation of trans complexes [10]. Reactions with [Pd(PPh3)4] afforded trans complexes. It is reported that metals belonging to the same group, such as Pt and Pd, have similar chemical properties. Studies on Pt and Pd complexes show the differences between Pd and Pt complexes are of kinetic nature [19,20]. Considering the similarity in the chemical properties of platinum and palladium ions, we compared Pt and Pd complexes assuming that the different behaviour would be mainly due to the difference in the cis or trans geometry of the complexes. [M(PPh3)4] precursors were used in the complexation reactions since it is reported that phosphine ligands by themselves and phosphine complexes of other metals, such as Ag(I), Au(I) and Sn(IV) are anticancer, anti-HIV or antimitochondrial agent [21–23]. Furthermore thymine and thymidine ligands were chosen since they are natural
255
components of DNA, and have been shown alter the pharmacological activity compound complexed with them. 2. Materials and methods Chemicals were obtained from commercial suppliers (Fluka) and used without further purification. The 1H and 13C NMR spectra were recorded on a Bruker WM400 spectrometer at 400 and 100.13 MHz respectively. All chemical shifts are expressed in ppm with respect to the signal of the protonated solvent (CD3OD: d 3.31 and 49.5; DMSO-d6: d 2.55 and 39.5). The 31P NMR spectra were run on a Bruker WM-400 spectrometer at 161.98 MHz, with external reference to 85% H3PO4 (d 0.0). The Fast Atom Bombardment (FAB) mass spectra (positive) were recorded on a ZAB 2SE spectrometer. The IR spectra were recorded on a Perkin–Elmer 457 spectrophotometer. Elemental analyses were performed by the ‘‘Analytische Laboratorien Prof. Dr. H. Malissa und G. Reuter GmbH’’ Industriepark Kaiserau (Haus Heidbruch) D-51789 Lindlar Germany. 2.1. Crystallography Intensity data collection was performed using graphite˚ ) and a monochromated Mo-Ka radiation (k = 0.70930 A pulse-high discrimination on an automated MACH3 Enraf-Nonius for complex 1. Intensity data collection for complex 2 was performed using graphite-monochromated ˚ ) and a pulse-high discrimCu-Ka radiation (k = 1.54178 A ination on a CAD4 Enraf-Nonius. Data analysis was performed as described in previous papers [17,18]. 2.2. Synthesis of complexes 1 and 2
Scheme 1.
Scheme 2.
1 equivalent of 1-methylthymine was reacted with 1 equivalent of [M(PPh3)4] (M = Pt, Pd) in refluxing dry toluene in the presence of KCl (1 equivalent). After stirring 5 h, toluene was evaporated. The crude mixture was redissolved in 7:3 v/v C6H6/CH3COOCH2CH3 and purified by silica gel chromatography. 1: Pd complex containing 1-methylthymine: Yield: 59%; Rf 0.4 (eluent C6H6/CH3COOCH2CH3 7/3 v/v). 1H NMR: (DMSO-d6) d 7.8–7.4 (15H, C6H5–P, m); 6.3 (1H, H6, bs); 2.6 (3H, CH3–N1, s); 1.2 (3H, CH3–C5, bs). 13C NMR: (CDCl3) d 170.1 (C4); 154.7 (C2); 134.8–127.7 (C6H5–P); 107.9 (C5); 36.1 (CH3–N1); 13.1 (CH3–C5). 31P NMR: (DMSO-d6) d 23.5. FAB-MS: (106Pd) 771m/z; [M + H]+. IR [(CHCl3) (m cm1)]: 1670, 1640 (C@O); 510 (PdAN). Molecular structure of trans-[PdCl(1-methylthymine) (PPh3)2] Æ H2O, 1 [18]. The palladium atom displays square planar coordination: two trans corners of the square plane are occupied by the phosphorus atoms of two triphenylphosphines. The chlorine atom and the amide nitrogen of the 1-methylthymine ligand are trans. Crystal data. C42H37ClN2O2P2Pd Æ H2O, MW = 822.01, monoclinic, space group name P21/c (no. 14),
256
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
˚ , b = 12.955(6) A ˚ , c = 19.190(10) A ˚, b= a = 17.780(10) A 3 ˚ 117.08(5)°, V = 3936(4) A , T = 298 K, Z = 4, l(Mo Ka) = 0.794 mm1, 7473 unique reflections (Rint = 0.0) used in all calculations. The final Rw(F2) was 0.1908; R1 = 0.0707. 2: Pt complex containing 1-methylthymine: Yield: 82%. Rf 0.44 (eluent CHCl3–MeOH 97:3, v/v). 1 H NMR (DMSO-d6): d 8.20–7.00 (30H, phenyl protons, m); 6.31 (1H, H-6, s); 3.0 (3H, CH3-1, s); and 1.62 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6): d 169.9 (C4); 155.1 (C-2); 138.6 (C-6); 109.3 (C-5); 36.1 (CH3-1); and 12.1 (CH3-5). 31P NMR (DMSO-d6): d 15.4 [d, PPh3 trans to Cl, 1J(Pt–P) = 3984] and 8.6 [d, PPh3 cis to Cl, 1 J(Pt–P) = 3235 Hz]. FAB MS (195Pt, 35Cl): m/z 893, [M + H]+ and 957 [M + Cl]+. IR [(CHCl3) (m cm1)]: 1690, 1640 (C@O). Molecular structure of cis-[PtCl(1-MeThy)(PPh3)2] Æ MeOH, 2 [17]. The platinum atom displays square planar coordination: two cis corners of the square plane are occupied by the P atoms of two triphenylphosphines and the chlorine atom and the amide nitrogen of the methylthymine ligand are in cis position to each other. The four atoms co-ordinated to the metal lie in a plane with very ˚ ) deviation from it. The plane of small (less than 0.08 A the methylthymine ligand is approximately perpendicular to the platinum coordination plane. Crystal data. C42H37ClN2O2P2Pt Æ CH3OH, M = 923.23, ˚, b= triclinic, space group P 1 (no. 2), a = 11.176(8) A ˚ , c = 13.36(1) A ˚ , a = 97.31(6)°, b = 91.32(6)°, 13.892(9) A c = 88.00(6)°, U = 2056(3) A3, T = 293 K, Z = 2,l(Cu-Ka) = 8.029 mm1, 7789 unique reflections (Rint = 0.0) used in all calculations. The final wR(F2) was 0.1334; R1 = 0.0515. 2.3. Synthesis of complexes 3, 4a and 4b Procedure described for 1 using as substrate. 3 0 ,5 0 -di-Oacetylthymidine. 3: Pd complex containing 3 0 ,5 0 -di-O-acetylthymidine: Yield: 87%. Rf 0.7 (eluent CHCl3/CH3OH 98:2 v/v). 1H NMR: (DMSO-d6) d 7.7–7.5 (15H, C6H5–P, m); 6.4 (1H, H6, bs); 5.8 (1H, H1 0 , dd); 5.1 (1H, H3 0 , m); 4.2 (2H, H5 0 , m); 4.1 (1H, H4 0 , m); 2.5 (2H, H2 0 , m); 2.1 (3H, CH3–CO, s); 2.0 (3H, CH3–CO, s); 1.2 (3H, CH3–C5, bs). 13C NMR: (CDCl3) d 169.8 and 170.1 (CH3–CO); 153.7 (C2); 168.9 (C4); 134.7–127.71 (C6H5–P); 84.2 (C1 0 ); 108.9 (C5); 80.7 (C3 0 ); 73.8 (C4 0 ); 63.4 (C5 0 ); 36.7 (C2 0 ); 20.4, 20.6 (CH3–CO); 13.4 (CH3–C5). 31P NMR: (DMSO-d6) d 23.5. FAB-MS: (106Pd) 991 m/z; [M + H]+. IR [(CHCl3) (m cm1)]: 1769, 1690, 1660 (C@O); 520 (PdAN). Anal. Calcd for C50H47ClN2O7P2Pd: C, 60.55; H, 4.78; Cl, 3.57; N, 2.82; P, 6.25; Found: C, 59.69; H, 4.60; Cl, 3.92; N, 3.10; P,7.03. 4a and 4b: Pt complexes containing 3 0 ,5 0 -di-O-acetylthymidine : Yield of the mixture: 80%. Rf 0.4 (eluent CHCl3–MeOH 97:3, v/v).
Complex 4a: 1H NMR (DMSO-d6) d 8.32–7.00 (30H, phenyl protons, m), 6.67 (1H, H-6, s), 6.17 (1H, H-1 0 , dd), 5.14 (1H, H-3 0 , m), 4.23 (2H, H-5 0 , m), 4.10 (1H, H4 0 , m), 2.14 and 2.07 (3H each, CH3CO, s), 2.1–1.9 (2H, H-2 0 , m) and 1.47 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6): d 170.0 (CH3CO), 167.9 (C-4), 153.6 (C-2), 134.7–127.5 (phenyl carbons and C-6), 110.1 (C-5), 83.6 (C-49), 80.4 (C-19), 73.2 (C-39), 63.6 (C-59), 35.4 (C-29), 20.7 and 20.5 (2CH3CO) and 13.3 (CH3-5). 31P NMR (DMSO-d6) d 15.3 [d, PPh3 trans to Cl, 1J(Pt–P) = 3960] and 8.6 [d, PPh3 cis to Cl, 1J(Pt–P) = 3265 Hz]. Complex 4b: 1H NMR (DMSO-d6) d 8.30–7.00 (30H, phenyl protons, m), 6.64 (1H, H-6, s), 6.26 (1H, H-1 0 , dd), 5.19 (1H, H-3 0 , m), 4.26 (2H, H-5 0 , m), 4.13 (1H, H-4 0 , m), 2.25 (2H, H-2 0 , m), 2.14 and 2.11 (3H each, CH3CO, s) and 1.48 (3H, CH3-5, s). 13C-{1H} NMR (DMSO-d6) d 170.0 (CH3CO), 167.8 (C-4), 154.1 (C-2), 134.7–127.5 (phenyl carbons and C-6), 109.2 (C-5), 83.6 (C-4 0 ), 80.3 (C-1 0 ), 74.2 (C-3 0 ), 63.6 (C-5 0 ), 35.3 (C-2 0 ), 20.7 and 20.5 (2 CH3CO) and 13.3 (CH3-5). 31P NMR (DMSO-d6) d 15.3 [d, PPh3 trans to Cl, 1J(Pt–P) = 3960] and 8.6 [d, PPh3 cis to Cl, 1J(Pt–P) = 3265 Hz]. FAB MS on mixture (195Pt, 35 Cl): m/z 1044, [M–Cl]+ and 754 [M-nucleoside]+. IR [(CHCl3) (m cm1)] on mixture 1750, 1663, 1590 [strong, m(CO)], 308 cm1 [weak, m(Pt–Cl)]. Anal. Calcd for C50H47ClN2O7P2Pt: C, 55.58; H, 4.38; Cl, 3.28; N, 2.59; P, 5.73; Found: C, 55.70; H, 4.42; Cl, 3.15; N, 2.30; P, 5.65. 2.4. Synthesis of complexes 5 and 6 0.1 g of derivative 3 (or 4) were dissolved in 2 mL CH3OH; 2 mL of NH4OH (37%) were added. The mixture was treated at 50 °C 2 h; the solvent was evaporated and the product was purified by silica gel column chromatography. 5: Pd complex containing thymidine: Yield:60%. Rf 0.8 (eluent 85:15 CHCl3/CH3OH v/v). 1 H NMR: (CD3OD) d 7.9–7.4 (15H, C6H5–P, m); 6.9 (1H, H6, bs); 5.7 (1H, H1 0 , dd); 4.2 (1H, H3 0 , m); 3.8 (1H, H4 0 , m); 3.7 (2H, H5 0 , m); 2.1 (1H, H2 0 , m); 1.7 (1H, H2 0 , m); 1.4 (3H, CH3-C5, bs). 13C NMR: (CD3OD) d 172.3 (C4); 156.1 (C2); 136.2–129.6 (C6H5–P); 110.2 (C5); 87.1 (C1 0 ); 88.5 (C4 0 ); 71.9 (C3 0 ); 63.1 (C5 0 ); 41.7 (C2 0 ); 13.9 (CH3-C5). 31P NMR (CD3OD): d 26.6 IR [CHCl3 (m cm1)]: 1667, 1646 (C@O); 520 (Pd–N). 6a and 6b mixture: Pt complexes containing thymidine: Yield: 90%. Rf 0.2 (eluent 85:15 CHCl3/CH3OH v/v). 1H NMR (CD3OD): d 7.80–7.10 (60H, phenyl protons and H-6, m), 6.19 and 6.01 (1H each, H-1 0 , dd), 4.33 (2H, H3 0 , m), 3.87 (2H, H-4 0 , m), 3.72 (4H, H-5 0 , m) 2.30–2.00 (4H, H-2 0 , m), 1.69 and 1.62 (3H each, CH3-5, s). 13C{1H} NMR (CD3OD): d 172.7 (C-4), 156.9 (C-2), 137– 128.0 (phenyl carbons and C-6), 111.9 and 108.7 (C-5), 88.8 (C-4 0 ), 87.5 and 86.9 (C-1 0 ), 72.5 and 72.4 (C-3 0 ), 63.2 (C-5 0 ), 41.8 and 41.4 (C-2 0 ) and 13.8 (CH3-5). 31P NMR (CD3OD): d 13.5 [d, PPh3 trans to Cl, 1J(Pt– P) = 3971] and 6.8 [d, PPh3 cis to Cl, 1J(Pt–P) = 3305
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
Hz]; IR [(CHCl3) (m cm1)] 1655, 1578 [strong, m(CO)], 300 cm1 [weak, m(Pt–Cl)]. 3. Results
257
Pt complexes 6a and 6b. Pt complexes 4a and 4b resulted more stable than Pd complex 5. Free thymidine nucleoside was isolated after deprotection of 5 by NH3. No cis–trans isomerization was observed in basic ammonia medium, as demonstrated by 31P NMR spectra.
3.1. Chemistry 3.2. Biological assays Complexes 1–4 were synthesized reacting [M(PPh3)4] (M = Pt and Pd) complexes and KCl respectively with N1-methylthymine and 3 0 ,5 0 -di-O-acetylthymidine in refluxing toluene, following procedures reported in the literature [17,18]. All reactions afforded M(II) metal complexes, characterized by an M-N3 of thymine or thymidine bond. Structure of the complexes was determined by NMR, IR, FAB-MS; X-ray data for complexes 1 and 2 and elemental analysis for complexes 3, 4a and 4b are reported. Reaction of [Pt(PPh3)4] with 3 0 ,5 0 -di-O-acetylthymidine afforded two diastereoisomeric species (4a and 4b), rapidly interconverting one into the other, whose presence can be ascribed both to the restricted rotation around the Pt–N3 bond and to the presence of fixed configurations at the sugar carbons [17]. When the same reaction was carried out with [Pd(PPh3)4] no diastereoisomers were isolated, because in this case the rotation around the Pd–N3 is very fast [18]. 1H and 31P NMR spectra of 3 recorded in CDCl3 at low temperature (15 °C) showed always just one signals pattern confirming no diastereoisomers can be isolated. All Pt complexes are characterized by the two triphenylphosphine in a cis geometry as demonstrated by the presence of two signals in the 31 P NMR, while Pd analogues have trans triphenylphosphines (just one signal in the 31P NMR spectrum) and the chloride ion trans to the base N3 (Schemes 1 and 2). Xray analysis performed on thymine complexes 1 and 2, revealed the presence of distorted square planar complexes. To further characterize Pd and Pt complexes, we performed ligand exchange experiments by NMR. Ligands are not displaced by an S donor such as DMSO as demonstrated by 1H and 31P NMR experiments in DMSO on complexes 3, 4a and 4b. In fact we did not observe any change in the ligand signals in the NMR spectra recorded at different times and scanning the temperature up to 50 °C. Furthermore treatment of 3 with a very reactive sulphorated molecule, such as CS2, did not cause ligand exchange, even after few days reaction [18]. More exchange experiments were carried out in excess of D2O (O donor) and of thymidine (N, O donor). We did not observe in the 1H and 31P NMR spectra neither a shift nor a variation in the integration for the signals of the ligands, which could have been due to the presence of displaced molecules. With the aim to obtain complexes soluble in aqueous media we deblocked the hydroxyl functions on deoxyribose moiety by ammonia treatment. This experiment was useful for testing the stability of the complexes to NH3 exchange. Treatment of complexes 3, 4a and 4b with NH3(aq) 37% at 50 °C gave complexes 5, 6a and 6b, in which the 3 0 and 5 0 position of the sugar have been deacetylated with yields ranging from 40% for the Pd complex 5 to 90% for the
Arrested PCR was performed as an assay for determining the biological activity in vitro [24,25]. This assay has been demonstrated useful for preliminary screening of DNA-binding drugs [24–26]. Briefly, target DNA was pre-incubated for 15 min with increasing concentrations of cisplatin or analyzed compounds and then PCR reaction was conducted with the 5 0 -TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-3 0 (forward) and 5 0 -CTA GAA GCA TTT GCG GTG GAC GAT GGA GGG-3 0 (reverse) primers, leading to an amplification of a 550 bp PCR product of b-actin, which was analyzed by agarose gel electrophoresis. PCR conditions were (20 cycles): denaturation 94 °C, 1 min; annealing 63 °C, 1 min; elongation: 72 °C, 1 min. The antiproliferative effect was analyzed on the human chronic myelogenous K562 cell line [27], obtained from American Type Culture Collection (ATCC, Rockville, MD) and grown in vitro in RPMI 1640 medium (Sigma/ Aldrich, Milwaukee, WI, USA), supplemented with 10% fetal bovine serum (FBS) (Celbio, Milano, Italy), 2 mM L-glutamine (Sigma/Aldrich, Milwaukee, WI, USA), a solution of 50 U/mL penicillin and 50 lg/mL streptomycin (Sigma/Aldrich, Milwaukee, WI, USA) in 5% CO2/air humidified atmosphere at 37 °C [28,29]. K562 cells were usually seeded at 30.000/mL and their treatment with compounds 1–6, triphenylphosphine (PPh3) and tryphenylphosphine oxide (PPh3O) was carried out by adding the appropriate drug concentrations at the beginning of the cultures. All the compounds were dissolved in MeOH. Control cultures containing the vehicle were always run in parallel to demonstrate absence of biological effects of MeOH under the experimental conditions employed. The medium was not changed during the induction period. To determine the antiproliferative activity of these compounds, exponentially growing cells were continuously exposed to various concentrations of drugs for 72 h and cell growth was studied by determining the cell number/ mL using Z1 Coulter Counter (Coulter Electronics, Hialeah, FL, USA) [29]. A representative example of arrested PCR data obtained is shown in Fig. 1 and the results from three different experiments summarized in Table 1, which suggests that important differences exist when the activities of the tested compounds are compared each other and with that of cisplatin. The most active Pd-compounds is compound 5, the most active Pt-compound is compound 6. When structurally related Pd- and Pt-compounds are compared (compare compound 3 with 4 and compound 5 with 6), it appears that Pd-compounds are more active than Ptcompounds in inhibiting PCR amplification.
258
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
Fig. 1. Effects of compounds 1–6, cisplatin, PPh3 and PPh3O on generation of PCR products. Target genomic DNA was pre-incubated for 15 min with increasing lM concentrations of compounds, as indicated and then PCR reaction was conducted with the 5 0 -TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-3 0 (forward) and 5 0 -CTA GAA GCA TTT GCG GTG GAC GAT GGA GGG-3 0 (reverse) primers. PCR products were analyzed by agarose gel electrophoresis and ethidium-bromide staining (M: marker). Table 1 Effects of compounds 1–6, triphenylphosphine and triphenylphosphine oxide on polymerase chain reaction (PCR) and cell growth of K562 cells Compound
Inhibition of PCR (IC50, lM)
Inhibition of cell growth (IC50, lM)
Cisplatin 1 2 3 4 5 6 Triphenylphosphine Triphenylphosphine oxide
8.5 ± 1.5 15.0 ± 5.8 25.5 ± 3.5 3.5 ± 0.8 7.0 ± 1.7 1.5 ± 0.3 3.1 ± 0.9 >50 >50
2.10 ± 0.91 4.25 ± 1.10 28.25 ± 8.20 2.25 ± 0.62 3.73 ± 0.81 0.58 ± 0.12 2.95 ± 1.11 5.07 ± 1.00 >100
A representative experiment on the effects on K562 cell growth is reported in Fig. 2 and the results of four independent experiments are shown in Table 1 and expressed as IC50 (dose causing 50% inhibition of cell growth in treated cultures relative to untreated controls). For each drug concentration, duplicate cultures were used. Vehicle or solvent controls were run in each experiment. The results on inhibition of K562 cell growth demonstrate that only one molecule is inactive (PPh3O). As found in arrested PCR experiments, Pd-compounds (among which the most active is compound 5) are more active than Pt-compounds. 4. Discussion The antiproliferative activity of cis and trans[M(PPh3)2LCl] complexes was tested on human
Fig. 2. Effects of compounds 1–6, PPh3 and PPh3O on K562 cell growth. K562 cells were cultured for 3 days in the presence of the indicated concentrations of compounds and then the cell number/ml was determined. d, compound 1; s, compound 2; j, compound 3; h, compound 4; m, compound 5; n, compound 6; r, PPh3; and e, PPh3O.
chronic myelogenous K562 cells. Based on early studies on bifunctional Pt cis and trans complexes it was postulated that the biological activity of metal complexes is due to their interaction with DNA; drug interstrand and intrastrand crosslinking cause lesion to DNA which inhibit its replication. It is now clear that cis and trans complexes interact with DNA in a different way; interstrand and intrastrand linkages are mainly mediated by guanine pairs. Square planar d8 Pt complexes have been shown to intercalate into DNA trough p–p and electrostatic interactions
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
[30], while Hoechst derivatives bind to the minor groove [31]. Monofunctional adduct, either cis and trans, are also able to bind DNA [32,33] and to terminate in vitro DNA and RNA synthesis [34]. Sterically hindered Pt complexes, such as [Pt(amine)2hetCl]+ (het: N-heterocyclic ligand) bind to DNA after one amine ligand, trans to the chloride has been lost [35]. In this work, planar aromatic ligands such as thymine or thymidine were used as metal ligands for their ability to interact with nucleic acids by stacking between two adjacent bases of a DNA strand. The presence of thymine should help the binding between the metal complex and the nucleic acids, introducing an additional binding site and directing the complex also toward adenine containing sequences. The steric hindrance of the ligands should slow down the reaction rate with the DNA and probably affect the reaction mechanism and the selectivity for the binding sites [35]. In addition it is reported that hindered complexes are used for circumventing resistance by sterically inhibiting cellular detoxification by glutathione and other cellular thiols [36]. The presence of at least one leaving group in the synthesized complexes should allow the reaction with DNA, either for the cis and the trans complexes, as demonstrated in the literature [35]. In previous papers we investigated the reactivity of [M(PPh3)] (M = Pd, Pt) complexes toward DNA bases, such as thymidine and thymine. Now we are interested in the characterization of the antiproliferative activity of the complexes containing t/T and Pd/Pt in cells. In order to investigate the stability of ligands in the described complexes we started studying the reactivity of our complexes in ligand exchange reactions. It is in fact known that ligand exchange reactions occurring in cells by thiols such as glutathione, inactivate metal complexes. S-donors ligands, such as DMSO, CS2, O-donors such as H2O, N,O-donors such as thymine, were used in exchange experiments. Reactions were followed by 1H and 31P NMR; the position and integration of the signals relative to bound ligands was not modified by ligands added, suggesting no exchange occurred. Based on these results we proceeded to the study of the antiproliferative activity of complexes in K562 cells. Analysis of the antiproliferative activity of Pt and Pd complexes reveals it is related to the hydrophilicity of the complexes: N1-methylthymine complexes 1 and 2 are less active than thymidine complexes 5, 6a and 6b. Complexed acetylated nucleosides (3 and 4) show the same IC50 no matter what the metal is, while for the corresponding free O-acetyl complexes 5 and 6 IC50 are significatively different, Pd complex resulting more active than Pt complex (Table 1). Reasonably these results can be explained considering the greater up-take of trans complexes; it is in fact reported that trans-Pt complexes are better internalized into cells than cis complexes [37]. Likely, DNA adducts formed by cis and trans compounds are different. The difference in the reaction rate of Pd and Pt complexes (Pd complexes are known for their lability in ligand exchange [15]), in this case should not affect the DNA com-
259
plexation rate. In order to determine whether the biological activity of the synthesized compounds was due to the entire complexes and not to lost ligands, the activity of phosphine ligands and phosphine oxide was tested; it is known that phosphine complexes lose ligands and that in oxidative conditions phosphines go to phosphine oxide. IC50 for phosphines was found 5.07 ± 0.95, while for phosphine oxide it was >100. Although in experimental NMR conditions the complexes resulted stable, it is not possible to exclude that exchange reaction occur in cells with release of phosphines. The IC50 measured for PPh3 results comparable to that of complexes 1, 3, 4 and 6, thus not allowing us to say whether the activity of the complexes can be attributed to the phosphines released or to the complexes as they are. With the exception of PPh3, arrested PCR data are strongly in agreement with the effects on cell growth, suggesting that differential inhibition of DNA replication by the tested compounds fully explains their in vitro antiproliferative activity. Additional mechanism(s) of action should be the molecular basis of the effects of PPh3 on K562 cell growth. 5. Conclusions Monofunctional, sterically hindered, cis and trans complexes [M(PPh3)2LCl] were synthesized by oxidative addition of [M(PPh3)4] to thymine and thymidine. The stability of the complexes was tested and confirmed by NMR ligands exchange experiments. All complexes were tested in antiproliferative assay on the human chronic myelogenous K562 cell line. Analyses of IC50 show that the antiproliferative activity: – is due to DNA binding as suggested by PCR inhibition, – is related to the hydrophilicity of the complexes, which probably affect their biodistribution and biodisponibility, – is strongly correlated with inhibition of PCR reactions and – is higher for the trans free O-acetyl complex 5. Furthermore, in the hypothesis all complexes 1–6 interact with DNA and ligand exchange does not occur as demonstrated by our in vitro experiments, it is likely that DNA lesions repair, which takes place in vivo by substitution of ligands with thiourea and glutathione, would not be observed because of the high stability of the complexes. Acknowledgment RG is funded by AIRC. References [1] D. Wang, S.J. Lippard, Nat. Rev. Drug Discov. 4 (2005) 307–320. [2] B. Rosenberg, L. VanCamp, J.E. Trosko, V.H. Mansour, Nature 222 (1969) 385–386.
260
A. Messere et al. / Journal of Inorganic Biochemistry 101 (2007) 254–260
[3] S.J. Lippard, Science 218 (1982) 1075–1082. [4] E. Monti, M. Gariboldi, A. Maiocchi, E. Marengo, C. Cassino, E. Gabano, D. Osella, J. Med. Chem. 48 (2005) 857–866. [5] W. Friebolin, G. Schilling, M. Zoller, E. Amtmann, J. Med. Chem. 47 (2004) 2256–2563. [6] L. Giovagnini, L. Ronconi, D. Aldinucci, D. Lorenzon, S. Sitran, D. Fregona, J. Med. Chem. 48 (2005) 1588–1595. [7] N. Bianchi, F. Ongaro, C. Chiarabelli, L. Gualandi, C. Mischiati, P. Bergamini, R. Gambari, Biochem. Pharmacol. 60 (2000) 31–40. [8] B. Rosenberg, L. Vancamp, T. Krigas, Nature 205 (1965) 698–699. [9] A.L. Pinto, S.J. Lippard, Biochim. Biophys. Acta 780 (1985) 167–180. [10] G. Natile, M. Coluccia, Coord. Chem. Rev. 216–217 (2001) 383–410. [11] N. Farrell, T.T. Ha, J.P. Souchard, F.L. Wimmer, S. Cros, N.P. Johnson, J. Med. Chem. 32 (1989) 2240–2241. [12] M.J. Clarke, F. Zhu, D.R. Frasca, Chem. Rev. 99 (1999) 2511–2534. [13] J. Kuduk-Jaworska, A. Puszko, M. Kubiak, M. Pelczynska, J. Inorg. Biochem. 98 (2004) 1447–1456. [14] G. Zhao, L. Huakuan, P. Yu, H. Sun, S. Zhu, X. Su, Y. Chen, J. Inorg. Biochem. 73 (1999) 145–149. [15] E. Budzisz, U. Krajewska, M. Rozalski, A. Szulawska, M. Czyz, B. Nawrot, Eur. J. Pharm. 502 (2004) 59–65. [16] V. Alverdi, L. Giovagnini, C. Marzano, R. Seraglia, F. Bettio, S. Sitran, R. Graziani, D. Fregona, J. Inorg. Biochem. 98 (2004) 1117– 1128. [17] L. De Napoli, R. Iacovino, A. Messere, D. Montesarchio, G. Piccialli, A. Romanelli, F. Ruffo, M. Saviano, J. Chem. Soc., Dalton Trans. (1999) 1945–1949. [18] A. Romanelli, R. Iacovino, G. Piccialli, F. Ruffo, L. De Napoli, C. Pedone, B. Di Blasio, A. Messere, Organometallics 24 (2005) 3401– 3406. [19] S. Gonzalez, J.M. Tersero, A. Matilla, J.M. Perez, V.M. Gonzalez, C. Alonso, J. Niclos-Gutirriez, J. Inorg. Biochem. 61 (1996) 261–272.
[20] U. Frey, S. Elmroth, B. Modlet, L.I. Elding, A.E. Merbach, Inorg. Chem. 30 (1991) 5033–5037. [21] S.J. Berners-Price, P.J. Dasler, Coord. Chem. Rev. 151 (1996) 1–40. [22] Y. Inouye, T. Kanamori, T. Yoshida, X. Bu, M. Shionoya, T. Koike, E. Rimura, Biol. Pharm. Bull. 17 (1994) 243–250. [23] H. Mitsuya, S. Broder, Nature 325 (1987) 773–778. [24] M. Passadore, N. Bianchi, G. Feriotto, C. Mischiati, P. Giacomini, R. Piva, R. Gambari, Biochem. J. 308 (1995) 513–519. [25] M. Passadore, N. Bianchi, G. Feriotto, C. Mischiati, C. Rutigliano, R. Gambari, Eur. J. Pharmacol. 319 (1997) 317–325. [26] P.G. Baraldi, B. Cacciari, R. Romagnoli, G. Spalluto, R. Gambari, N. Bianchi, M. Passadore, P. Ambrosino, N. Mongelli, P. Cozzi, C. Geroni, Anticancer Drug Des. 12 (1997) 555–576. [27] B.B. Lozzio, C.B. Lozzio, Int. J. Cancer 19 (1977) 136. [28] E. Fibach, N. Bianchi, M. Borgatti, E. Prus, R. Gambari, Blood 102 (2003) 1276–1281. [29] P.G. Baraldi, R. Romagnoli, N. Bianchi, R. Gambari, Bioorg. Med. Chem. 11 (2003) 2381–2388. [30] E.R. Jamieson, S.J. Lippard, Chem. Rev. 99 (1999) 2467246–2467298. [31] M.A. Carrondo, M. Coll, J. Aymami, A.H. Wang, G.A. van der Marel, J.H. van Boom, A. Rich, Biochemistry 28 (1989) 7849–7859. [32] V. Brabec, V. Boudny, Z. Balcarova, Biochemistry 33 (1994) 1316– 1322. [33] M.B. Janik, B. Lippert, J. Bio. Inorg. Chem. 4 (1999) 645–653. [34] Z. Balcarova, J. Kasparkova, A. Zakovska, O. Novakova, M.F. Sivo, G. Natile, V. Brabec, Mol. Pharmacol. 53 (1998) 846–855. [35] M.B. Janik, A. Hegmans, E. Freisinger, B. Lippert, J. Bio. Inorg. Chem. 4 (1999) 130–139. [36] T. Okada, I.M. El-Mehasseb, M. Kodaka, T. Tomohiro, K. Okamoto, H. Okuno, J. Med. Chem. 44 (2001) 4661–4667. [37] N. Farrell, L.R. Kelland, J.D. Roberts, M. Van Beusichem, Cancer Res. 52 (1992) 5065–5072.