Introduction

Top Summary Introduction Experimental Procedures Results Discussion References

It has been postulated that the DNA interstrand cross-link adducts formed by platinum drugs may contribute to the cytotoxicity of these compounds (Leng and Brabec, 1994). In fact, the interstrand cross-links formed in DNA by cis-diamminedichloroplatinum(II) (cis-DDP) between guanine residues in opposite strands preferentially inhibit DNA or RNA polymerases in vitro (Lemaire et al., 1991; Vrána et al., 1996). Moreover, the binding of platinum drugs on DNA may induce conformational changes of the double helix structure that may be responsible for the drug activity (Johnson et al., 1989; Leng and Brabec, 1994; Pil and Lippard, 1997). The secondary structure most commonly adopted by DNA in vivo is the B-form, which is right-handed; however, DNA may also adopt other conformations (Cantor, 1981). Thus, a new conformational state of the DNA, namely Z-DNA, characterized by a left-handed conformation of the helix, has been described (Pohl and Jovin, 1972; Herbert and Rich, 1996). It is now well established that DNA sequences that are in Z conformation or having potential to adopt the Z-DNA form can be found in biological systems (Nordheim et al., 1981; Nordheim and Rich, 1983; Lancilloti et al., 1987; Casasnovas et al., 1989; Wittig et al., 1991). Moreover, a possible role for Z-DNA in gene regulation has also been proposed (Jiménez-Ruíz et al., 1991).

Pohl and Jovin showed that at high salt concentration, poly(dG-dC)·poly(dG-dC) undergoes a cooperative, conformational transition from the B-DNA form toward the Z-DNA form (Pohl and Jovin, 1972). Later on, Behe and Felsenfeld noticed that the methylated polymer poly(dG-m⁵ dC)·poly(dG-m⁵dC) also undergoes a transition from the B form to the Z form, but it does so at lower salt concentrations (Behe and Felsenfeld, 1981). Malinge and Leng (1984) found that binding of the antitumor drug cis-DDP to poly(dG-m⁵dC)·poly(dG-m⁵dC) in B conformation induces a conformational change toward a distorted Z-like form and that binding of cis-DDP to the polynucleotide in Z conformation stabilizes its Z-DNA structure. Moreover, it has been also observed that a family of trans-bis(platinum) complexes provokes the B-to-Z transition in poly(dG-m⁵dC)·poly(dG-m⁵dC) (Johnson et al., 1992). In contrast, the clinically ineffective trans-isomer of cis-DDP inhibits B-to-Z transition in DNA (Peticolas and Thomas, 1985; Rahmouni et al., 1985; Zaludová et al., 1997).

We have previously reported that the cytotoxic bis(platinum) complex [Pt₂Cl₄(diminazene aceturate)₂]Cl₄·4H₂O (Pt-berenil) (Fig. 1) induces drastic changes on DNA secondary structure, leading to a compaction of the double helix, and that the diminazene aceturate (berenil) ligand directs the DNA binding of the Pt-berenil drug (González et al., 1996, 1997). We show herein that the interaction between the Pt-berenil compound and DNA results in the formation of a large number of covalent interstrand cross-links and that this number is significantly higher than that caused by cis-DDP-DNA interaction. Our data indicate, moreover, that the kinetics of DNA interstrand cross-link formation by Pt-berenil binding is faster than that of cis-DDP-DNA interaction and that the number of DNA interstrand cross-links in Pt-berenil-DNA complexes is influenced by supercoiling. To determine whether the large number of DNA interstrand cross-links formed by the drug affect the conformational changes of the double helix, we have analyzed the effect of Pt-berenil binding on the MgCl₂-induced B-DNA-to-Z-DNA transition of poly(dG-m⁵dC)·poly(dG-m⁵dC). We have chosen this methylated polymer because the dinucleotide sequence "m⁵dC-dG" appears so frequently in eukaryotic DNA [it may account for more than half of all "d(CpG)" sequences (Razin and Riggs, 1980)]. Moreover, we have chosen MgCl₂ as the Z-DNA inducer because it allows work at Cl concentrations close to those found intracellularly (Rich et al., 1984). The results show that Pt-berenil strongly inhibits the salt-induced B-DNA-to-Z-DNA transition of poly(dG-m⁵ dC)·poly(dG-m⁵dC). We think that these results are interesting in view of the fact that the currently known platinum drugs, in contrast with Pt-berenil, induce the B-DNA-to-Z-DNA transition (Malinge and Leng, 1984; Johnson et al., 1992; Zaludová et al., 1997). Because we also present data here demonstrating that Pt-berenil is active in cis-DDP-resistant cells, we propose that other bis(platinum) compounds with structures similar to Pt-berenil might be designed in the search for new cytotoxic platinum compounds capable of circumventing cisplatin resistance.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Chemical structure of Pt-berenil (A) and berenil (B).

	Experimental Procedures

Top Summary Introduction Experimental Procedures Results Discussion References

Materials. The stock solutions of the cis-DDP, berenil, and Pt-berenil compounds were prepared to a final concentration of 1 mg/ml. Pt-berenil was synthesized as described previously (González et al., 1996). Cis-DDP was kindly supplied by Bristol-Myers S.A. (Madrid, Spain). The Pt-berenil and cis-DDP compounds were dissolved in 10 mM NaClO₄. The drug solutions were freshly prepared before use.

Cytotoxicity Assays. The ovarian carcinoma cell lines A2780, A2780cisR, CH1, CH1cisR, SKOV-3 and SKOV-3cisR were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% of newborn calf serum together with 2 mM glutamine, 100 units/mL penicillin, and 100 mg/mL streptomycin at 37°C in an atmosphere of 95% air/5% CO₂. A2780cisR, CH1cisR, and SKOV-3cisR are a sublines of their respective parent lines that have acquired cisplatin resistance (Kelland et al., 1993). Cell survival in compound-treated cultures was analyzed by using a system based on the tetrazolium compound 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide, which is reduced by living cells to yield a soluble formazan product that can be assayed colorimetrically (Alley et al., 1988). Cells were plated in 96-well sterile plates, at a density of 10⁴ cells/well in 100 µl of medium and were incubated for 3-4 h. Compounds dissolved in DMEM were added to final concentrations from 0.1 to 100 µM, in a volume of 100 µl/well. Ninety-six hours later, 50 µl of a freshly diluted 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide solution (1/5 in culture medium) was added at a concentration of 1 mg/ml into each well and the plates were incubated for 5 h at 37°C in a humidified 5% CO₂ atmosphere. Cell survival was evaluated by measuring the absorbance at 520 nm, using a Whittaker Microplate Reader 2001. All experiments were made in quadruplicate.

Platination Reactions. Calf thymus DNA, pF18 DNA, and poly(dG-m⁵ dC)·poly(dG-m⁵dC) at concentrations of 200 µg/ml were incubated with the platinum drugs at r_i = 0.1 in 10 mM NaClO₄ at 37°C. At various time periods (10 min, 30 min, 1 h, 2 h, 3 h, 18 h, and 24 h), aliquots of the reaction mixture were withdrawn and assayed by total X-ray fluorescence (TXRF) for platinum not bound to DNA (Niemann et al., 1990; Wobrauschek, 1994; González et al., 1996). The molar ratio of platinum bound to nucleotides (r_b) was calculated by subtracting the amount of free (unbound) platinum present in the reaction mixture. In agreement with previously reported data (González et al., 1996), it was observed that the kinetics of DNA platination by both Pt-berenil and cis-DDP were similar and that after 24 h of incubation, DNA platination reached a plateau because > 95% of the cis-Pt(II) centers from both Pt-berenil and cis-DDP were bound to DNA.

DNA Renaturation Experiments. Sonicated calf thymus DNA (average length of 500 base pairs) was dissolved in 0.02 × standard saline citrate (150 mM NaCl, 15 mM sodium citrate, pH 7.0) at a concentration of 20 µg/ml. Aliquots of 1 ml of the calf thymus DNA solution were modified by the drugs at r_b of 0.05. The DNA was melted at 45°C to 95°C. After 2 min at 95°C, control DNA was allowed to reanneal by decreasing the temperature from 95°C to 12°C above the melting temperature of native DNA (57 ± 0.5°C) at a 1°C/min rate. DNA modified by Pt-berenil or cis-DDP were exposed to comparable renaturation conditions. Thus, Pt-berenil-modified DNA or cis-DDP-modified DNA were allowed to reanneal by decreasing the temperature at a 1°C/min rate from 95°C to 12°C above the melting temperature of Pt-berenil-DNA (62 ±0.7°C) or cis-DDP-DNA complexes (55 ± 0.3), respectively. The decrease in chromicity upon renaturation was measured in a Beckman Acta CIII spectrophotometer attached to a temperature programmer. The percentage of DNA renaturation was calculated from the hypochromicity value at 95°C.

Interstrand Cross-Link Assay. To linearize the pF18 plasmid, the DNA was digested in 150 mM NaCl with 10 units/µg DNA of BamHI (unique restriction site in pF18) at 37°C for 4 h. The plasmid DNA was 3'-end labeled by incubation with 2.5 mCi/mg DNA of [-³²P]dCTP and 1.25 units/µg DNA of the Klenow fragment of E coli DNA polymerase I for 30 min at room temperature. The reaction was stopped by heating at 70°C for 5 min. The unincorporated radioactivity was removed by passing the labeling reaction through a Sephadex G-50 column. The labeled DNA was precipitated with 0.1 volumes of sodium acetate and 2 volumes of cold ethanol. Sonicated DNA was added to the eluted solution of the labeled pF18 DNA to a final DNA concentration of 180 µg/ml. Afterwards, the DNA, at a concentration of 90 ng/ml, was incubated with the Pt-berenil and cis-DDP drugs in 10 mM NaClO₄ at r_i of 0.1 for different periods of incubation. Then, 10-µl aliquots were removed and the reaction was ended by addition of an equal volume of the loading dye (90% formamide, 10 mM EDTA, 0.1% xylene cyanol, and 0.1% bromphenol blue). At each period of incubation, the Pt bound to DNA (r_b) was monitored by TXRF. The DNA was melted for 10 min at 90°C and chilled on ice. Agarose gel (1.5%) electrophoresis in denaturing conditions was carried out at 20 V for 16 h (Maniatis et al., 1989). Interstrand cross-link formation was also detected in covalently closed circular pF18 plasmid DNA (density of supercoiling,  = 0.067). Supercoiled pF18 plasmid was incubated with Pt-berenil or cis-DDP under the conditions indicated above and subsequently linearized with BamHI endonuclease, 3'-end labeled with [-³²P]dCTP, denatured by heat, and subjected to agarose gel electrophoresis in denaturing conditions. The gels were dried and autoradiographed. Band quantification was made using a Molecular Dynamics model 300A densitometer (Sunnyvale, CA).

Determination of the Cross-Linking Rate. The number of cross-links was estimated assuming a Poisson distribution [(XL/fg = ln(1  F_DS)] (Vos and Hanawalt, 1987). XL/fg represents the number of cross-links per fragment. F_DS results from dividing the peak of the integrated area of the double-stranded band by the sum of the integrated areas of the double-stranded and single-stranded bands. The (XL/fg)/fragment size ratio gives the number of cross-links per kilobase pair (kb).

Electron Microscopy of Pt-berenil-pF18 DNA Complexes. Supercoiled ( = 0.067) or linear pF18 plasmid DNA (2 µg) were incubated at 37°C in 10 mM NaClO₄ with Pt-berenil at r_i = 0.1 until an r_b = 0.05 was achieved. Then, the DNA was precipitated with 2.5 volumes of cold ethanol and 0.1 volume of 3 M sodium acetate, pH 4.8. Afterward, the precipitate was washed with 100 µl of 75% ethanol. Sample aliquots of the Pt-berenil-pF18 DNA complexes (DNA concentration = 0.1 µg/ml) were denatured and extended as a monolayer on an aqueous hypophase according to the method of Sogo and Thoma (1989). The DNA interstrand cross-links were visualized in electron micrographs of the denatured supercoiled and linear DNA samples.

Determination of Platinum Bound to DNA In Vivo. Culture plates containing 10 ml of HeLa cells (human cervix carcinoma line) in DMEM (cell density, 2 × 10⁵ cells/ml) were preincubated at 37°C for 24 h in an atmosphere of 5% CO₂. Then, 500 µl of culture medium containing the desired amount of drug to reach a final drug concentration of 35 µM (the IC₅₀ value of Pt-berenil in HeLa cells) was added to the plates. Subsequently, the plates were incubated for 48 h at 37°C and 5% CO₂. Afterward, the cells were centrifuged at 1500 rpm for 5 min, washed with phosphate-buffered saline, resuspended in 400 µl of buffer containing 20 mM Tris·HCl, pH 7.5, 2 mM EDTA, and 0.4% Triton X-100, incubated at 4°C for 15 min, and centrifuged at 12,000 rpm during 15 min in a Microfuge (Hettich Microfuge, Madrid, Spain). The supernatants were then treated for 3 h at 37°C with 20 mg/ml of proteinase K in a buffer containing 150 mM NaCl, 40 mM Tris·HCl, pH 8.0, 40 mM EDTA, and 1% SDS. Finally, the DNA was extracted with phenol/chloroform/isoamyl alcohol (25:24:1), precipitated with 2.5 volumes of cold ethanol and 0.1 volumes of 3 M sodium acetate, washed with 75% ethanol, dried, and resuspended in 1 ml of water. The DNA content in each sample was measured by UV spectrophotometry at 260 nm. The platinum bound to DNA was determined by TXRF (Niemann et al., 1990; Wobrauschek, 1994). Experiments were carried out in triplicate.

In Vivo Detection of DNA Modifications Induced by Binding of Pt-Berenil. The HeLa cells cultivated and treated with 35 µM concentrations of the Pt-berenil drug for 24 h were harvested with trypsin/EDTA and centrifuged. The pellets were treated as described in the previous section to isolate the DNA. Aliquots (250 µl) of genomic DNA (60 ng/ml of DNA) were digested to completion with EcoRI. Subsequently, 20-µl aliquots were subjected to 1% agarose gel electrophoresis during 16 h at 1.5 V/cm and transferred to a nylon membrane (Amersham). Southern blot detection of genomic DNA was performed using a [³²P]-labeled 18 S ribosomal DNA probe (10 mCi/ml).

CD Spectroscopy. CD spectra were recorded in a JASCO J-600 spectropolarimeter interfaced to a 486 PC. Measurements were performed at 37°C using 1-cm path-length cells. Each spectrum represents the mean of three scans. Aliquots containing 15 µg/ml of poly(dG-m⁵ dC)·poly(dG-m⁵dC) were prepared from the stock solution. CD spectra were run in a range of wavelength from 220 to 320 nm and at a speed of 50 nm/min. Scans were recorded at 0.4 nm intervals.

	Results

Top Summary Introduction Experimental Procedures Results Discussion References

Cytotoxicity of Pt-Berenil in Cisplatin-Resistant Cells. We reported previously that Pt-berenil has a cytotoxic activity in a micromolar range similar to that of cisplatin against HL-60 and U-937 leukemic cells (González et al., 1996). In view of these encouraging results, we have tested the cytotoxicity of Pt-berenil against an ovarian carcinoma panel, including cells sensitive and resistant to cis-DDP. Table 1 shows the IC₅₀ values of Pt-berenil and cis-DDP against A2780, A2780cisR, CH1, CH1cisR, SKOV, and SKOVcisR tumor cells (Kelland et al., 1993). It may be observed that Pt-berenil has a cytotoxic activity similar to that of cis-DDP in the A2780, CH1, and SOV cisplatin-sensitive cells (IC₅₀ values for Pt-berenil of 0.20 µM, 0.70 µM, and 0.30 µM, respectively, versus IC₅₀ values for cis-DDP of 0.30 µM, 3.30 µM, and 0.10 µM, respectively). Interestingly, however, in A2780cisR, CH1cisR, and SKOVcisR cells, Pt-berenil circumvents the acquired cisplatin resistance because this compound exhibits IC₅₀ values of 0.70 µM, 0.36 µM, and 10.40 µM, respectively, versus IC₅₀ values for cis-DDP of 3.30 µM, 0.65 µM, and 39.60 µM, respectively. Thus, the data suggest that the Pt-berenil complex may be considered to be a potential antitumor agent, because it has resistance factor indexes (see Table 1) that are significantly lower than cisplatin against the panel of human ovarian carcinoma cells tested.

DNA Interstrand Cross-Link Formation by Pt-Berenil. Previously reported data indicated that the binding of Pt-berenil to DNA significantly increases the melting temperature of the double helix above that produced by the ligand berenil and that the increase is incubation time-dependent (González et al., 1996). DNA renaturation experiments of Pt-berenil-DNA and berenil-DNA complexes suggested that Pt-berenil might induce covalent DNA interstrand cross-links because the percentage of DNA renaturation in Pt-berenil-DNA was much higher than that of berenil-DNA. The rationale of the experiments was that linked homologous sequences should have, in solution, renaturation kinetics faster than the same nonlinked sequences because of a "zippering" effect provided by a nucleation site for helix formation at the cross-link site. Figure 2 shows that the percentage and kinetics of reassociation of the DNA of the Pt-berenil-DNA complexes is significantly higher than that of the control DNA and that of the berenil-DNA complexes. Because after 25 min of reassociation, the chromicity of the Pt-berenil-DNA complexes was the same as that of the native Pt-berenil-DNA complexes, it is likely that most of the DNA of the complex had renatured and that interstrand nucleation points had been formed along the DNA. That the nucleation sites of the DNA from the Pt-berenil-DNA complexes might be caused by stable interstrand cross-links may be deduced from the fact that the bridging of berenil to DNA after denaturation of the helix does not affect the reassociation kinetics. The incubation time of the DNA with the Pt-berenil drug influences the number of the nucleation points, because the percentage of DNA reassociation is proportional to the period of complex formation (data not shown). A similar assay was performed with the drug cis-DDP, which has been shown to form interstrand cross-links at a low ratio (Lemaire et al., 1991). The results obtained confirmed previous reported data indicating that the number of interstrand cross-links (potential nucleation sites) caused by cis-DDP-DNA interaction should be low, because the percentage and kinetics of DNA reassociation relative to native DNA was not significantly altered.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Kinetics of DNA renaturation of control DNA (), Pt-berenil-DNA (), cis-DDP-DNA (), and berenil-DNA () complexes formed at rb = 0.05.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   A, pattern of single and double stranded DNA bands after melting of the cis-DDP-pF18 DNA, Pt-berenil-pF18 DNA and berenil-pF18 DNA complexes. The complexes were formed by incubation of the drugs with the DNA at an ri value of 0.1 for 10 min, 30 min, 1 h, 2 h, and 3 h. (ds, double stranded form; ss, single stranded forms; Cn, native DNA; Cd, denaturated DNA; B, berenil). B, cross-linking rates for cis-DDP-DNA () and Pt-berenil-DNA (). The mean rb values obtained for Pt-berenil at the incubation times indicated above were 0.005, 0.010, 0.015, 0.030, and 0.048, respectively, and for cis-DDP were 0.002, 0.010, 0.017, 0.036, and 0.047, respectively. The standard deviation in all cases was <5%.

View larger version (59K): [in this window] [in a new window]   Fig. 4.   A, pattern of single- and double-stranded DNA bands after BamHI digestion and subsequent melting of the cis-DDP-pF18 DNA and Pt-berenil-pF18 DNA formed in covalently closed circular pF18 DNA ( = 0.067) at ri values of 0.1 for 10 min, 30 min, 1 h, 2 h, and 3 h. Lane 1, control unmelted pF18 DNA; lane 2, control melted pF18 DNA; lanes 3 through 7, cis-DDP-pF18 DNA complexes formed after 10 min, 30 min, 1 h, 2 h, and 3 h, respectively; lanes 8 through 12, Pt-berenil-pF18 DNA complexes formed after 10 min, 30 min, 1 h, 2 h, and 3 h, respectively. (dsDNA, double-stranded pF18 DNA; ss, single-stranded pF18 DNA). The rb mean values obtained in supercoiled pF18 DNA for Pt-berenil at the incubation times indicated above were 0.007, 0.014, 0.020, 0.040, and 0.055, respectively, and for cis-DDP were 0.005, 0.016, 0.021, 0.038, and 0.052, respectively. The standard deviation in all cases was <5%.

View larger version (140K): [in this window] [in a new window]   Fig. 5.   Electron micrographs of linear pF18 DNA (A), supercoiled pF18 DNA (B) and linear (C) and supercoiled (D) pF18 DNA incubated with Pt-berenil at rb = 0.05. The supercoiling density, , of covalently closed circular pF18 DNA was 0.067.

Binding of Pt-Berenil to DNA In Vivo. The binding of Pt-berenil and cis-DDP molecules to the DNA in whole cells was determined by quantification of Pt atoms bound to DNA using TXRF (Niemann et al., 1990; Wobrauschek, 1994). The data indicate that under the conditions used, the number of Pt-berenil molecules bound to genomic DNA in Pt-berenil treated HeLa cells was 3-fold higher than the number of molecules bound to the genomic DNA in the cis-DDP treated cells. Thus, at a concentration of 35 µM, the drug, 50 µg of DNA, and 0.016% Pt-berenil was bound to DNA, compared with 0.005% of cis-DDP. We observed, moreover, that the lower DNA binding rate of cis-DDP was associated with its slower cell uptake relative to Pt-berenil since after 48 h of incubation of the Hela cells with 35 µM, the drugs (equivalent to 2.1 × 10¹⁷ drug molecules), 93% of the input Pt-berenil molecules and 84% of the input cis-DDP molecules, were transported into the cells (data not shown).

View larger version (38K): [in this window] [in a new window]   Fig. 6.   Hybridization of a 18 S ribosomal DNA probe to genomic DNA from HeLa cells digested with EcoRI endonuclease. Lane 1, DNA from nontreated HeLa cells. Lane 2, DNA from HeLa cells treated with 35 µM Pt-berenil.

Effect of Pt-Berenil on the MgCl₂-Induced B-DNA to Z-DNA Transition of Poly(dG-m⁵ dC)·Poly(dG-m⁵dC). Table 2 shows the MgCl₂ concentrations and the times needed to reach the B-to-Z transition midpoint (t_1/2) of control polynucleotide and of the polynucleotide modified by Pt-berenil and berenil at r_b = 0.05. It may be observed that at an MgCl₂ concentration of 1.3 mM, the B-DNA-to-Z-DNA transition of poly(dG-m⁵dC)·poly(dG-m⁵dC) has a midpoint-time of 65 min. Similar results were obtained when the polynucleotide was modified by berenil, although this drug provoked a delay of the B-to-Z transition, because the midpoint-time was obtained after 87 min of incubation with the 1.3 mM MgCl₂. Surprisingly, however, when the polynucleotide was modified by Pt-berenil, the B-to-Z transition midpoint was not reached even after 120 min at the same MgCl₂ concentration. Moreover, it is interesting to note that after 5 min of incubation with 3 mM MgCl₂, the shape of the CD spectrum of the polynucleotide modified by Pt-berenil is characteristic of the B conformation. In contrast, an inversion of the CD spectrum characteristic of the Z conformation is observed in the polynucleotide modified by berenil and in control polynucleotide after 5 min of incubation with 3 mM MgCl₂ (Fig. 7).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   CD spectra of poly(dG-m5dC)·poly(dG-m5dC) (), poly(dG-m5dC)·poly(dG-m5dC)-Pt-berenil complexes formed at rb = 0.05 (... ... ...) and poly(dG-m5dC)·poly(dG-m5dC)-berenil complexes (---) formed at rb = 0.05 after 5 min of incubation with 3 mM MgCl2 at 37°C. Ellipticity units are given in mdeg. (... ... ...) Ellipticity baseline (at 0).

	Discussion

Top Summary Introduction Experimental Procedures Results Discussion References

We have presented in this article evidence showing the ability of the cytotoxic compound Pt-berenil to form drug-DNA interstrand cross-links. The kinetics and percentage of reassociation of the DNA from the Pt-berenil-DNA complexes suggested that the renaturation of the helix is caused by the existence of a "zippering" effect. Most probably, Pt-DNA interstrand cross-link adducts were formed at the "zippering site." This interpretation is consistent with other data showing that several platinum compounds are also able to form DNA interstrand adducts (Lemaire et al., 1991). When we compared the percentage of renaturation of the DNA from the Pt-berenil-DNA complexes with that of the DNA from cis-DDP-DNA complexes we observed that the renaturation rate was higher in the Pt-berenil treated DNA than in the cis-DDP-treated DNA, as an indication that the number of "zippering" points related to interstrand cross-links should be high in the Pt-berenil-treated DNA.

The existence of interstrand cross-links in Pt-berenil-treated DNA was shown by the presence of double-stranded DNA forms in plasmids treated with the drug after denaturation. The data revealed that at r_b around 0.05, there is at least one interstrand cross-link per linear pF18 plasmid DNA molecule and that, on the average, the number of DNA interstrand cross-links produced by Pt-berenil is 11 times higher than that of cis-DDP. Moreover, the data indicate that under our experimental conditions, the number of interstrand cross-links per plasmid molecule produced by Pt-berenil in supercoiled pF18 DNA is about 2-fold higher than that produced by the drug in linear DNA (3.75 versus 1.98). On the other hand, cis-DDP also increases the number of interstrand cross-links in supercoiled pF18 DNA relative to linear pF18 DNA, but to a much lower extent than Pt-berenil. These results are in agreement with previously reported data indicating that supercoiling affects the formation of interstrand adducts by cis-DDP (Vrána et al., 1996). A similar phenomenon has been also observed in the binding to DNA of several other drugs, such as psoralen or ethidium bromide (Cook et al., 1989).

Electron micrographs taken from preparations of denatured Pt-berenil-pF18 DNA complexes confirmed that the intercatenary interaction of the Pt-berenil drug with the DNA through the cis-Pt(II) centers is favoured in supercoiled plasmids, whereas in linear pF18, only one site of cross-link was present in most of the DNA molecules several sites of cross-linking were visualized in the supercoiled DNA forms. Thus, because Pt-berenil unwinds the double helix in supercoiled plasmids (González et al., 1996), it is expected to bind more readily to negatively supercoiled DNA.

The CD data show that binding of Pt-berenil to poly(dG-m⁵dC)·poly(dG-m⁵dC) produces a strong inhibition of the salt-induced B-DNA-to-Z-DNA transition. In fact, when the polynucleotide is modified by the drug at r_b = 0.05, the B-to-Z equilibrium is not reached under salt conditions (3 mM MgCl₂) at which the native polynucleotide readily adopts the Z-DNA form (Fig. 7; Behe and Felsenfeld, 1981; Gonzalez et al., 1998). Interestingly, the CD data indicate that the polynucleotide modified by Pt-berenil is in B form even after 5 min of incubation with this salt concentration. On the other hand, our CD data indicate that the inhibition of the B-DNA-to-Z-DNA transition of poly(dG-⁵dC)·poly(dG-m⁵dC) produced by berenil is much lower than that of Pt-berenil. Because Pt-berenil forms a high number of DNA interstrand cross-links through the cis-Pt(II) centers ]in contrast with berenil, which binds electrostatically to DNA (Brown et al., 1990)[, it is most likely that such a high number of interstrand cross-links may block the uncoiling of the DNA double helix required to undergo the B-DNA-to-Z-DNA transition.

The cytotoxicity data reported in this article indicate that Pt-berenil circumvents resistance to cisplatin in several ovarian tumor cell lines. It has been postulated that the DNA interstrand cross-links formed by platinum drugs may contribute to the antitumor activity of these compounds (Leng and Brabec, 1994). Therefore, it is tempting to speculate about the possibility that the ability of Pt-berenil to form DNA interstrand cross-links may be in part responsible for its cytotoxic activity in cis-DDP-resistant cells.

In summary, the results presented here support the hypothesis that the synthesis of platinum compounds having biologically relevant molecules as ligands might lead to the development of cytotoxic agents active in cisplatin resistant cells and with potential to form specific types of DNA adducts.