Antitumour Activity of a pt(III) Derivative of 2-Mercaptopyrimidine

The complex [Pt2Cl2(Spym)4], where Spym = 2-mercaptopyrimidine, was synthesized and analyzed spectroscopically. The presence in the 195Pt NMR spectrum, of only one signal for the Pt(III) indicates the symmetrical arrangement of the ligands and the identical setting of N, S and Cl atoms, PtS2ClN2, for the two Pt atoms being different to other compounds described in the literature. The interaction of this complex with DNA was studied by several techniques, including circular dichroism, melting temperature determination, electron microscopy (EM) and atomic force microscopy (TMAFM). Preliminary results show a high activity against HL-60 and HeLa tumour lines for the Pt-2-mercaptopyrimidine complex in comparison with cisplatin activity. Higher values for IC50 were obtained, while the values of LD50 were lower than those for cisplatin.


Introduction
2-Mercaptopyrimidine is a potential model of biologically relevant molecules such as 2-thiouracil and 2-thiocosine that are sometimes present in t-RNA, and it has also been tested as bacterial growth inhibitor and antiviral agent(') However, until now, it has not been used as a drug.
The development of the pyrimidine blues, poiymeric compounds in which mixed-valence metal ions, usually platinum, are bridged by ligands derived from pyrimidines or amides, and the discovery of their high antitumour activity have extended the study to pyrimidine related-Pt(lll) complexes. Several Pt(lll) dimers have [been described (4) and the structure of a dimer of Pt(lll) with 2mercaptopyrimidine was determined by Goodgame et a/. (5).
There are several possibilities of coordination for the 2-mercaptopyrimidine molecule ( Figure 1 ). It can bind to the metal as a neutral ligand via the S atom (A), or in bidentate fashion via N and S atoms (B), or as an anionic ligand (C), or finally as a bridging ligand to two metal atoms (D).   Goodgame(5) (Figure 2). Here we present a complete spectroscopic study together with a description of its effect on DNA structure and the results of several biological tests.

Materials and Methods
The complex Pt-Spym was prepared using K2[PtCI4] Johnson Matthey and 2mercaptopyrimidine Aldrich. Elemental analyses were carried out on a Carlo Erba 1500 microanalyzer at the Serveis Cientffico-Tcnics at the University of Barcelona. The infrared spectra were recorded in solid star? (KBr pellets) on an FT-IR Nicolet 5DZ spectrometer in the 4000-400 cm " range. H{3C}, 3C{H} and 95pt{H}NMR spectra were obtained on a Brucker DRX 250 sp.ectrometer using DMSO-d6 as solvent. Chemical shifts were measured relative to TMS in the case of H and 3C and to Na2PtCI6 for 95pt. UV spectra were recorded at room temperature on a double beam SHIMAZDU UV-2101-PC spectrometer. The absorbance change during DNA denaturalization (Tin) was measured on the same SHIMAZDU UV-2101-PC spectrometer with thermostatic cells connected to an automatic temperature regulator NESLAB RTE-110 bath. The heating rate was 1.5 9C/min. The absorbance was measured at , 260 nm. CD spectra were obtained on a JASCO J720 spectropolarimeter with a 450 W xenon lamp. Samples, prepared as described below, were observed on transmission electronic microscopes Philips EM 200 and Philips EM 301 at 80,000 V, the first working at 25,800 and 54,900 magnification and the second at 16,000, 20,000 and 26,000 magnification. Over 82.5 ml TAE, 25 ml 0.1% cytochrome c, 25 ml 38% formaldehyde and 610 ml pure water solution, the adduct PBR322:Pt-Spym with ri 0.50 and 72 h of incubation was added. Microdrops were prepared from this solution and after 20 rain Ni MESH200 grills recovered with FORMVAR membrane and reinforced with graphite were deposited over DNA drops. The samples were shaded by evaporation of WO3 in an evaporation chamber at 5.10 .6 torr, under nitrogen atmosphere for 10 rain by a 20 A and 150 W current with a rotation of 100 rpm.
Images of linear DNA, 200 base-pair copy of Escherichia coli, and adducts of Pt compounds and this DNA were obtained with Extended Nanoscope III (Digital Instruments, Santa Barbara, CA) working in TMAFM mode in about 100 nN. The samples were prepared as described above and 2 ml of each was deposited on a small mica disk, washed in distilled water after 20 rain, and dried under argon.
In vitro studies, cytotoxicity in tumour lines and type of cell death were determined in Centro de Biologia Molecular "Severo Ochoa" at Universidad Aut(noma de Madrid-CSIC. Synthesis of the Complex 1 mmol of K2PtCI4 was dissolved in 10 ml of water and mixed with a suspension of 1 mmol of 2-mercaptopyrimidine (Spym) in 10 ml of hot water. After a few minutes a pale brown precipitate was formed and after stirring for 3 h at 40 C and 48 h at room temperature a dark brown solid appeared.
The product was filtered off the orange solution, washed with H20 and ethanol and finally dried over silica gel. Calculated for [Pt2CI2(Spym)4] C, 21 When the synthesis was repeated using the ratio K2PtCI4 Spym 1:2, the same product was obtained while the solution remained colourless.

RESULTS AND DISCUSSION FT-IR Spectra
The IR data for 2-merca. ptopyrimidine and theplatinum complex are shown in Table 1 The band at 1217 cmwas assigned to Vst(SH) of the thiol form, and the band t 1188 cm-.was assigned to v(NH) and v(C=S) of the thiona form.  These results indicate that the two tautomeric thiol and thiona forms are present in the solid (3,10-12).
Three weak bands at 2063, 1977 and 1915 cm " assigned to aromatic ring combination bands disappeared when the complex was formed due to the electronic changes induced by the coordination of the metal ion. The band at 1188 cm -1 assigned to v(C=S) of the thiona form ( 1,(13)(14) in 2-mercaptopyrimidine decreased in frequency and intensity in the complex indicating the coordination through the S atom (5). In contrast, the band at 1217 cm-1 corresponding to v(C-S) of the thiol form increased in frequency when the metal ion bound to 2-mercaptopyrimidine. The other bands related with S bonds also showed changes in frequency and intensity. Special attention should be given to the thioamide bands resultin.q in the coupling between different vibrations when a C=S group is bound to one or more N atoms (5). The band assigned to thioamide III (v(C=N) + v(C=S) at 983 cm -1 split and increased in frequency in the complex [Pt2CI2(Spym)4], as has been observed for bidentate (N,S) compounds. On the other hand, the Vst(SH) bands disappeared, indicating that the thiol form is not p_resent when the metal ion is coordinated by sulphur. A new band corresponding to v(Pt-S)( 8, [16][17][18] was also observed for the complex. The splitting of the band assigned to v(C=C) + v(C=N) at 1570 cmin the Pt compound indicates that the pyrimidine ring may be involved in the coordination to the metal. The appearance of a new band at 225 cm-1 assigned to 5(M-N) (8,9, 2o) and the behaviour of the thioamide III band seem to reinforce this hypothesis. A new band at 296 cmcan also be observed. Although the bands corresondin.q to terminal v(Pt-CI) in Pt(ll) compounds appeared at frequencies between 345 and 310 cm-l(9,21,22), the Pt(lll) dimers with terminal CI showed the v(Pt-CI) absorption at lower frequencies (23,24). This shifting is due to the lengthening of the Pt-CI bond caused by the trans effect of the Pt-Pt bond of the dimer (5,(25)(26)(27) (25). Although the complex can react with DMSO and the Pt-Pt bond can be broken easily to give monomeric species, no solvolysis signal was obseved. The low-temperature (246K) spectrum is the same as that recorded at room temperature.   In the complex, the equivalence between C(4) and C(6) signals is lost. The most noticeable shift corresponds to C(5), followed by C(4) and C(6). The quaternary C(2) appears as a very weak signal showing only an insignificant shift due to the low sensitivity of carbon atoms of this kind. The large shift of C(5) can be explained because this atom is in para with respect to the S atom coordinated to the pt (29). The N coordinated to the metal atom is probably the closest to C(4) because the corresponding signal shifts more than that assi.qned to C(6).
Finally, the 195pt spectrum for the complex [Pt2CI2(pym)4], shows a single signal at -1179 ppm, related to K2PtCI6, which represents an intermediate shift between those of Pt(ll) and Pt(IV) for a similar environment, and which corresponds to an oxidation state of Pt(111) (27,30,31).
The presence of only one signal for the Pt(lll) indicates the symmetrical arrangement of the ligands and the identical setting of N, S and CI atoms PtS2CIN2, for the two Pt atoms. In the case of the analogous compound described by Goodgame( )the crystal structure indicates that two types of Pt environment, PtSCIN3 and PtS3CIN maybe r)resent. The different chemical environment must lead to the appearance of two signals in the 195pt spectrum but the authors did not reported anything about it. However, the ecLui.valent dimer compound with iodine [Pt212(Spym)4], also studied crystallographically by Goodgameua, presents the same chemical environment PtS21N 2 for both Pt.

Circular Dichroism
The circular dichroism spectra of Caff Thymus DNA and Caff Thymus DNA incubated with the mercaptopyrimidine-Pt compound at several times and with several molar ratios were registered.
In Tables 4a, 4b   The major changes occur for ri 0.25. There was a slight modification of DNA spectrum when the Pt-Spym complex was bound to the helix. This change was more marked at 16 h of incubation. The positive band has less ellipticity than the DNA one and this effect, together with the increase of the ellipticity and the decrease of the wavelength, indicate changes in the base stacking, probably due to the opening of the helix(32) The behaviour is very similar to that induced by transplatin, which originates helix opening due t'o the binding between the two strains of the same DNA helix. The structure of the Pt-Spym compound that presents two labile sites (CI) at the ends of the dimer can allow a similar type of binding.  However, after 48 h of incubation, the changes are less marked, although the greatest alteration occurs also for ri 0. 25

UV. Melting Temperature (Tm)
The wavelengths and the absorbances corresponding to the maximum UV of the complex DNA:Pt(Spym) are collected in Table 5. In the adduct DNA:Pt-Spym a strong shifting of the maximum towards lower wavelenghts was observed. This effect is greater for higher ri values (0.25, 0.50) and a progressive increase in the hyperchromicity when the Pt/nucleotide ratio increased was also observed. This seems to indicate a strong modification of the secondary DNA structure. Going from ri 0.25 to ri 0.50 a Antitumor Activity of a Pt(III) Derivative of 2-Mercaptopyrimidine b Figure 3. Microphotographs corresponding to plasmid alone (a), and to plasmid incubated for 24 h. with Pt-Spym at ri 0.50 (b)  The compound Pt-Spym induces a slight decrease in Tm. In compounds having intercalator aromatic rings the values found for ATm are in the range 0.4-1 C for CT-DNA at pH= 7.0 (33).
On the other hand, this compound caused an important decrease in the absorbance at high temperatures.
These results together with the UV spectrum indicated a marked interaction with the DNA at room temperature.

Electron Microscopy
In Figure 3, the microphotographs corresponding to plasmid alone(a), to plasmid incubated with cisp/atin at ri 0.50 for 24 h (b), and to plasmid incubated with Pt-Spym at ri 0.50 for 24 h (c) are shown. The plasmid alone is in circular closed form with different degrees of supercoiling, and a low percentage of linear form can also be seen. The cisplatin causes a comlactation in the plasmid, as has been described in the literarure (34) which can reach up to 50 %. The Pt-Spym compound also causes a compactation, but also lateral aggregation with other plasmid molecules. This compactation is more intense than that of cisp/atin in the same experimental conditions. This fact may indicate that the Pt-Spym compound induces some kind of interaction between different plasmid molecules, probably due to the binding of platinum with two strains from two different molecules.
Atomic Force Microscopy (TMAFM) Figure 4 shows the images corresponding to linear fragment DNA, hliM (a), the same DNA incubated 24 h with cisp/atin (b), incubated 24 h with Pt-2-mercaptopyrimidine (c), and 24 h with 2mercaptopyrimidine alone (d). The concentration in DNA is the same in all the samples. Cisplatin seems to compact and distort. In the magnification corresponding to 400 nm, several points where the fragment is doubled can be observed. The measurement of the length of this fragment on the microscope reveals this modification. The measurement of the area also shows that there is no aggregation between different linear DNA fragments. So, the binding type of cisplatin to DNA, mainly to two N intra-strain sites, is reflected. However, after DNA incubation with Pt-Spym, always in the same conditions as for cisp/atin, aggregation and compacting phenomema were observed. The aggregate formed have larger area than a single fragment, indicating that the presence of the complex has produced the aggregation of several DNA fragments. So, the binding of Pt-Spym induces an interaction between different DNA strains. Probably, it binds through the two lateral Sites after Pt-CI has been hydrolyzed. The effect observed is not due to the ligand. The effect on the DNA in the same conditions as for 2-mercaptopyrimidine alone can be observed in Figure 4(c). The measurement of the length of the strain on the microscope confirms that neither lengthening or distorsion has been caused.

Preliminary Cytotoxicity tests on Tumour lines
The tests have been carried out on two tumour lines: HeLa cells from an uterus cancer and HL-60 cells from human leukemia. The trials were performed with cisplatin as reference, and with the Pt-2-mercaptopyrimidine complex. The LC50 parameters are summarized in Tables 7 and 8.
The values for the Pt-Spym are better than for the cisp/atin against the tumour line HL-60. The percentage of cell death is high at very low concentrations of complex, 3.5 mM, which is markedly lower than the one obtained for cisp/atin.  The difference between cisplatin and the Pt-Spym compound is still greater for the tumour line HeLA. Although the LC50 value for Pt-Spym is the same, cell death is practically complete at 5 mM. For cisplatin, the same effect was produced only at concentrations near 150 mM. 2-Mercaptopyrimidine alone did not produce cell death. Probably the dramatic effects of the platinum complex over DNA are related with the structure of the complex but the possible role of the oxidation state III of the platinum in the mechanisms of cytotoxicity cannot be ruled out (35). The analysis of the type of cell death was carried out with two kinds of tests: the observation of the characteristic morphological changes by means of phase contrast microscopy and the checking of DNA digestion in regular fragments by gel electrophoresis. This exposure in HeLA cells, at 16 or 24 h, produced the characteristic changes due to apoptosis.
The same test was carried out with the tumour line HL-60. The Pt-Spym does not induce apoptosis in this tumour line.
In vivo toxicity test with male mouses BDF gave a DL50 value for the Pt-Spym compound of 250 mg/Kg. This value indicates that the toxicity is much lower than those of cisplatin or carboplatin, which are the first and second generation platinum complexes used in clinical trials.