The resolution of 5,5,7-trimethyl-1,4-diazacycloheptane: coordination to platinum(II), crystal structure of the dichloroplatinum(II) complex and HPLC separation of the [Pt(GpG)(tmdz)] adducts

The resolution of 5,5,7-trimethyl-1,4-diazacycloheptane: coordination to platinum(II), crystal structure of the dichloroplatinum(II) complex and HPLC separation of the [Pt(GpG)(tmdz)] adducts

Polyhedron 18 (1999) 1039–1043 The resolution of 5,5,7-trimethyl-1,4-diazacycloheptane: coordination to platinum(II), crystal structure of the dichlo...

134KB Sizes 1 Downloads 36 Views

Polyhedron 18 (1999) 1039–1043

The resolution of 5,5,7-trimethyl-1,4-diazacycloheptane: coordination to platinum(II), crystal structure of the dichloroplatinum(II) complex and HPLC separation of the [Pt(GpG)(tmdz)] adducts 1 V.P. Munk, R.R. Fenton*, T.W. Hambley

School of Chemistry, University of Sydney, Sydney, NSW 2006, Australia Received 18 May 1998; accepted 20 October 1998

Abstract The resolution of the cyclic diamine tmdz (tmdz55,5,7-trimethyl-1,4-diazacycloheptane) is described. The absolute configuration was determined by preparation and crystal structure analysis of [PtCl 2 (R-tmdz)] by X-ray diffraction methods. Each enantiomer reacts with d(GpG) to form two isomers that were readily separated by HPLC. Significant stereoselectivity was observed in the reaction of the S enantiomer with d(GpG) but not in the reaction of the R enantiomer.  1999 Elsevier Science Ltd. All rights reserved. Keywords: Cyclc diamine tmdz; Crystal structure analysis; X-ray diffraction

1. Introduction Cisplatin hcis-[PtCl 2 (NH 3 ) 2 ], cis-diamminedichloroplatinum(II)j forms a variety of DNA adducts including intrastrand and interstrand adducts but which of these are responsible for the anticancer properties of the complex has yet to be determined [1–4]. The predominance of the intrastrand adducts is consistent with their being responsible but recent evidence suggests that the interstrand adduct may contribute to the anticancer activity of cisplatin [5–7]. We have been addressing these questions by designing compounds to form a subset of the adducts. For example, [PtCl 2 (hpip)] complex (hpip5homopiperazine51,4diazacycloheptane) was designed to be suited to forming interstrand adducts, but less suited to forming intrastrand adducts. This design goal was achieved, as reported elsewhere [8,9]. The reaction between [PtCl 2 (hpip)] and DNA was found to be stereoselective [8,10] in that two isomers formed in a ratio of approximately 1:3. The isomers differed in the orientations of the ethylene and propylene chains of the ligand with respect to the exocyclic oxygen atoms of the guanine bases [9,10]. These

results suggested that steric interactions between the complex and DNA can exert a strong influence on the adducts formed between Pt complexes and DNA [9]. To investigate this further we have turned to analogues of hpip with additional steric bulk such as the tmdz ligand (tmdz5 5,5,7-trimethyl-1,4-diazacycloheptane, Scheme 1, chiral centre indicated with an *) that are expected to be more stereoselective in their interactions with DNA. We have previously reported the preparation and crystal structure of racemic [PtCl 2 (tmdz)] and its reactions with d(GpG) [11]. However, since both tmdz and d(GpG) are chiral, diastereomeric products are expected and were observed [11]. Separation of all isomers was not possible and therefore we have undertaken resolution of the tmdz ligand. In order to understand the interactions between the enantiomers of [PtCl 2 (tmdz)] and DNA we have determined its absolute configuration and have investigated the reactions of each enantiomer with d(GpG).

*Corresponding author. 1 Corresponding co-author. 0277-5387 / 99 / $ – see front matter PII: S0277-5387( 98 )00390-8

Scheme 1. 

1999 Elsevier Science Ltd. All rights reserved.

1040

V.P. Munk et al. / Polyhedron 18 (1999) 1039 – 1043

2. Experimental

2.1. Instrumentation 1

H and 13 C NMR studies were performed on a Bruker AC 200F spectrometer, using commercially available solvents referenced to TMS or solvent isotopic impurities. All polarimetry experiments were carried out using an Optical Activity POLAAR 2001 Dual Wavelength Automatic Polarimeter, using a 1 dm cell and recorded at ambient temperature. Circular dichroism measurements were performed using a JASCO J-710 spectropolarimeter equipped with J-700 software for Windows. All spectra were recorded with a sensitivity setting of 200 mdeg and a spectral band width of 1.0 nm. Measurements were recorded between the wavelengths 260 nm and 600 nm. High-performance liquid chromatography (HPLC) studies of the [Pthd(GpG)j(tmdz)] complexes were carried out using a Series 800 HRLC system V2301a, equipped with a BIO-RAD UV-1806 detector, a Model AS-100 HRLC automatic sampling system and a Model 2800 solvent delivery system. The analytical column used was a Supercosil TM LC-18 column (5 mm particle size, 4.63250 mm).

2.2. Resolution of 5,5,7 -trimethyl-1,4 -diazacycloheptane The synthesis of 5,5,7-trimethyl-1,4-diazacycloheptane (tmdz) was carried out by a method adapted from Curtis [12] as described by Teo et al. [11]. The resolution of the racemic ligand was carried out using 2,3-dibenzoyl-tartaric acid. An aqueous solution of tmdz (9.8 g, 0.07 mol, 40 ml) was slowly added to a solution of (2)-2,3-dibenzoyl-Ltartaric acid (Aldrich, 24.73 g, 0.07 mol) in the minimum amount of ethanol (40 ml). The reaction mixture was gently boiled, and a white precipitate immediately formed upon cooling. This precipitate was collected and redissolved in boiling ethanol (50 ml). After cooling to room temperature the white crystals of S-tmdz-dibenzoyl-L-tartrate were collected, filtered at the pump, washed with the minimum amount of ice-cold water and air-dried. The S-tmdz-dibenzoyl-tartrate salt was recrystallised in the minimum volume of a boiling water / ethanol mixture (1:1) (|10 ml g 21 ), and the precipitate was collected as described above. The optical rotation of the S-tmdz-dibenzoyl-L-tartrate salt was measured in methanol at the sodium D line (589 nm). After several recrystallisations, optical rotation measurements were constant. Yield: 10.3 g, 30%, [a ] D ; 275.4860.5 (c51, methanol). Resolution of the other enantiomer, R-tmdz, was achieved by the same method as for S-tmdz, using (1)2,3-dibenzoyl-D-tartaric acid as the resolving agent. The dibenzoyl-tartrate salt was repeatedly recrystallised until constant rotation was achieved. Yield: 7.05 g, 20%, [a ] D ; 174.1860.5 (c51, methanol). To liberate the tmdz ligands from their dibenzoyl-

tartrate salts, the optically pure tmdz-dibenzoyl-tartrate salts were suspended in water (50 ml), cooled to 08C, and 10 M NaOH was slowly added until the pH was |14. The optically pure diamines were extracted into dichloromethane (4350 ml), dried over anhydrous sodium sulphate and the solvent removed by rotary evaporation resulting in a clear, brown oil. Optical rotations of both enantiomers were measured, and compared. [a ] D ; R-tmdz, 28.7860.5; S-tmdz, 19.4860.5 (c51, methanol). The 1 H NMR spectra of the free tmdz ligands in deuterated chloroform show identical peaks for the R- and S-enantiomers. Identification of the peaks is as follows, ppm (d ); 1.09 (m, 9H); 1.62 (m, 4H); 2.81 (m, 3H); 2.98 (m, 2H). These peaks are consistent with those observed for racemic tmdz [11].

2.3. Synthesis of enantiomerically pure [ PtCl2 (tmdz)] The synthetic precursor, cis-[PtCl 2 (DMSO) 2 ], was synthesised using a procedure outlined by Price et al. [13]. Enantiomerically pure tmdz (0.142 g, 1.00 mmol) in methanol (20 ml) was slowly added to a solution of cis-[PtCl 2 (DMSO) 2 ] (0.425 g, 1.00 mmol) suspended in methanol (40 ml). The pale yellow solution was stirred for 2 h, until there was no evidence of any solid in the solution. The solvent was then removed by rotary evaporation, taking care that the temperature remained below 408C. An excess of lithium chloride in water (20 ml) was slowly added and the mixture was gently heated until a solid began to form. The mixture was left to stand at room temperature overnight, and the pale yellow crystals were collected at the pump, washed with ice-cold water, absolute ethanol, diethyl ether and then air-dried. Yield: [PtCl 2 (R-tmdz)] 0.298 g, 0.73 mmol, 73%; [PtCl 2 (S-tmdz)] 0.240 g, 0.59 mmol, 59%. CD spectra showing equal but opposite absorption profiles have been deposited as an accessory publication. The 1 H NMR studies of the enantiomers of [Pt(tmdz)Cl 2 ] were carried out in D 2 O, using TPS (3(trimethylsilyl)-1-propane-sulphonic acid) as an internal standard (peaks d |0 ppm and d |4.8 ppm). The spectra of the R- and S-tmdz platinum(II) complexes are identical with resonances that are seen at; 1.08 (d, 3H); 1.23 (s, 3H); 1.63 (m, 2H); 1.84 (m, 3H); 2.90 (m, 1H); 3.20 (m, 2H); 3.85 (m, 2H), ppm. Identical 13 C NMR spectra of the enantiomers of [Pt(tmdz)Cl 2 ] were obtained in DMF–d 7 . Solvent peaks appear as two septets at d |30 ppm and d |34 ppm, and as a triplet at d |162 ppm. These peaks, due to isotopic impurities, were used as an internal reference. Only seven peaks were seen in the 13 C spectra, but it was confirmed by Distortionless Enhancement of Polarisation Transfer (DEPT) experiments that one peak is hidden beneath a solvent peak. Thus the eight expected carbon environments are seen at; 19.1 (s, CH 3 ); 26.2 (s, CH 3 ); 30.0 (s, CH 3 ); 40.3 (s, CH 2 ); 47.2 (s, CH 2 ); 49.7 (s, CH 2 ); 54.1 (s, CH); 59.4 (s, C), ppm.

V.P. Munk et al. / Polyhedron 18 (1999) 1039 – 1043

1041

Table 1 ˚ Bond lengths (A) Pt(1)–Cl(1) Pt(1)–N(1) N(1)–C(1) N(2)–C(3) C(1)–C(2) C(1)–C(7) C(3)–C(8)

2.310(3) 2.043(7) 1.52(1) 1.49(1) 1.53(1) 1.53(1) 1.56(1)

Pt(1)–Cl(2) Pt(1)–N(2) N(1)–C(5) N(2)–C(4) C(1)–C(6) C(2)–C(3) C(4)–C(5)

2.311(3) 2.018(8) 1.51(1) 1.50(1) 1.54(1) 1.50(1) 1.53(1)

2.4. X-ray crystallography Crystals suitable for crystallographic analysis were obtained by slow evaporation of a D 2 O solution in an NMR tube. Cell constants were determined by a leastsquares fit to the setting parameters of 25 independent reflections, measured and refined on an Enraf–Nonius CAD4-F diffractometer. Data reduction and application of Lorentz, polarisation and analytical absorption corrections were carried out using teXsan [14]. The structure was solved by direct methods using SHELXS-86 [15] and refined using full-matrix least-squares methods with teXsan [14]. Hydrogen atoms were included at calculated sites with thermal parameters derived from the parent atoms. Non-hydrogen atoms were refined anisotropically. Neutral atom scattering factors and anomalous dispersion terms were taken from International Tables [16]. Anomalous dispersion effects were included in Fc [17]; the values for Df 9 and Df 0 were those of Creagh and McAuley [18]. The values for the mass attenuation coefficients are those of Creagh and Hubbell [19]. All other calculations were performed using the teXsan [14] crystallographic software package of Molecular Structure Corporation. Bond lengths and angles are listed in Tables 1 and 2. The atomic nomenclature is defined in Fig. 1 [20]. Listings of atom coordinates, tables of torsion angles, anisotropic thermal parameters, details of least-squares planes calculations and observed and calculated structure factor amplitudes have been deposited.

Table 2 Bond angles (8) Cl(1)–Pt(1)–Cl(2) Cl(1)–Pt(1)–N(2) Cl(2)–Pt(1)–N(2) Pt(1)–N(1)–C(1) C(1)–N(1)–C(5) Pt(1)–N(2)–C(4) N(1)–C(1)–C(2) N(1)–C(1)–C(7) C(2)–C(1)–C(7) C(1)–C(2)–C(3) N(2)–C(3)–C(8) N(2)–C(4)–C(5)

93.3(1) 172.4(2) 94.2(2) 113.8(5) 111.9(7) 106.2(6) 110.2(7) 108.3(7) 109.6(8) 116.5(8) 113.0(7) 108.2(7)

Cl(1)–Pt(1)–N(1) Cl(2)–Pt(1)–N(1) N(1)–Pt(1)–N(2) Pt(1)–N(1)–C(5) Pt(1)–N(2)–C(3) C(3)–N(2)–C(4) N(1)–C(1)–C(6) C(2)–C(1)–C(6) C(6)–C(1)–C(7) N(2)–C(3)–C(2) C(2)–C(3)–C(8) N(1)–C(5)–C(4)

95.7(2) 170.9(2) 76.8(3) 103.4(5) 110.0(5) 113.6(8) 106.7(8) 111.5(7) 110.5(7) 112.6(7) 109.1(8) 110.5(7)

Fig. 1. ORTEP plot of complex molecule giving the crystallographic atom numbering. 30% probability ellipsoids are shown.

2.5. Crystal data Formula [PtCl 2 (C 8 H 18 N 2 )].H 2 O; Mr 426.25, orthorhombic, space group P2 1 2 1 2 1 , a 8.039(3), b 11.404(2), c ˚ V 1257.2(6) A ˚ 3 , Z 4, Dcalc 2.252 g cm 23 , 13.713(4) A, 21 ˚ F (000) m (MoKa) 115.14 cm , l(MoKa) 0.71069 A, 808 electrons, umax 25.0, T max 0.562, T min 0.116, N 1319, No 1195 (I. 2.5 s (I)), Nvar 127, max shift 0.001 s ), R 2 0.022, R w 0.018 and w51 /s 2 , Four max 0.7, Four min 21.0.

2.6. HPLC analysis of the adducts formed between the enantiomers of [ PtCl2 (tmdz)] and d( GpG) [Pthd(GpG)j(tmdz)] complexes were synthesised using a method outlined by Teo et al. [11] modified from previously described procedures [21–23]. The HPLC elutions were performed at a flow rate of 1 ml min 21 , with methanol in ammonium acetate (0.1 M, pH 5.5). The absorbance profiles were monitored at 254 nm and are shown in Fig. 2. Peak intensities were estimated by integration of the absorbance profiles using BIO-RAD software.

3. Results and discussion

3.1. Resolution of tmdz The resolution of racemic tmdz involved the separation

2

R5S(uuF o u2uF c uu) / SuF o u, R w 5S(w 1 / 2 uuF o u2uF c uu) / Sw 1 / 2 uF o u.

1042

V.P. Munk et al. / Polyhedron 18 (1999) 1039 – 1043

confirming the absolute configuration at C(3) as R as shown in Fig. 1. We have previously determined and reported the structure of racemic-[PtCl 2 (tmdz)] and therein encountered problems with spurious peaks in the final maps that interfered with the refinement and were not resolved by collecting a data set on a new crystal [11]. The present structure suffers from none of these problems and therefore represents a substantial improvement over the previous determination. Bond lengths are normal but bond angles reflect the steric strain induced by formation of the chelate rings. This is also reflected in the near zero torsion angle [2(1)8] for the N–C–C–N moiety. There are no ˚ from the least squares deviations greater than 0.025 A plane through the coordination plane. The ligand forms two chelate rings; the six-membered chelate ring adopts a chair conformation and the five-membered chelate ring an envelope conformation. Overall, the ligand conformation can be described as a boat with the two planar moieties meeting at an angle 113.88. The constrained ligand structure results in a close contact between the methyl group disposed toward the metal [C(6)] and the metal itself. The ˚ respecPt???H and Pt???C separations are 2.73 and 3.19 A tively, in close agreement with the values predicted using our recently revised force field for Pt(II) complexes of ˚ [24]. 2.66 and 3.15 A

3.3. HPLC analysis of the adducts formed between the enantiomers of [ PtCl2 (tmdz)] and d( GpG) Fig. 2. HPLC chromatograms for the reactions of (a) [PtCl 2 (R-tmdz)] with d(GpG) and (b) [PtCl 2 (S-tmdz)] with d(GpG).

The reaction between racemic [PtCl 2 (tmdz)] and d(GpG) was expected to yield the four isomers shown in Scheme 2 but HPLC analysis of the products revealed only three peaks [11]. Attempts to resolve these peaks further

of the two enantiomers, based on the differing solubilities of the diastereomeric (1,2)-2,3-dibenzoyl-D-tartaric acid salts in a water / ethanol solution. The successful isolation of the optically pure enantiomers was confirmed by recrystallisation to a constant rotation of the tartrate salts and by the observation that the rotations are equal in magnitude and opposite in sign. The magnitudes of the [a ] D values for the free ligand are similar but, as is often the case, the hygroscopic nature of diamines leads to inaccuracies in the [a ] D values and therefore, those of the tartrate salts are a more accurate indication of the successful resolution of the ligand. The enantiomeric purity of the platinum complexes was confirmed by measuring their circular dichroism spectra and, as expected, these revealed equal and opposite curves (Supplementary Fig. 1).

3.2. Description of the structure The structure consists of the neutral complex and one water molecule of crystallisation with hydrogen bonds between the water molecule and both H(amine) atoms and one of the chloro ligands. Inversion of the structure and re-refinement gave an R value of 0.038 (R w 0.036) thus

Scheme 2.

V.P. Munk et al. / Polyhedron 18 (1999) 1039 – 1043

by changing the chromatographic conditions were unsuccessful and this prompted us to resolve the tmdz ligand. Reaction of each enantiomer separately with d(GpG) is expected to yield two isomers (Scheme 2) and HPLC traces reveal two peaks in each case (Fig. 2). It is now clear that both enantiomers give rise to an isomer that elutes at 22.3 min and consequently resolution of the tmdz ligand has allowed us to resolve these four isomers. We are now in a position to characterise these isomers and investigate the reaction of each enantiomer with DNA and this work is proceeding. An estimation of the amount of each isomer observed was obtained by integrating the area of the HPLC peaks. For the R enantiomer the isomer that eluted at 21.2 min accounted for 54% and that at 22.3 min, 46%. This difference is marginal and is probably not significant given the approximate method used for estimating the peak area. However, there is a more substantial difference in the amounts of the isomers generated by the S enantiomer and this is visually apparent in the HPLC trace shown in Fig. 2 and in all other traces for this reaction. The isomer that eluted at 22.3 min accounts for 62% and that at 24.4 min for 38%. These differences, both within and between the two pairs of isomers, indicate significant steric interactions between the tmdz ligand and the incoming d(GpG) dinucleotide. No such difference was observed in the isomers generated by the reaction of [PtCl 2 (hpip)] and DNA [9,10]. Presumably then, the additional methyl groups present in the tmdz ligand do, as anticipated, provide a steric impediment to binding to nucleotides and thus we can expect to see an even more substantial difference in the reactions with DNA.

Acknowledgements The support of the Australian Research Council is gratefully acknowledged. Supplementary material available Tables of crystal data, positional and thermal parameters, bond lengths and angles and least-squares planes are available on request from the CCDC, 12 Onion Road, Cambridge, C62 IE7, UK.

1043

Circular dichroism spectra are available on request from the authors.

References [1] S.J. Lippard, Pure Appl. Chem. 59 (1987) 731. [2] J.L. van der Veer, J. Reedijk, Chemistry in Britain 24 (1988) 775. [3] N. Sheibani, M.M. Jennerwein, A. Eastman, Biochemistry 28 (1989) 3120. [4] A.M.J. Fichtinger-Schepman, P.H.M. Lohman, J. Reedijk, Nucleic Acids Res. 10 (1982) 5345. [5] J.C. Jones, W. Zhen, E. Reed, R.J. Parker, A. Sancar, V.A. Bohr, J. Biol. Chem. 266 (1991) 7101. [6] W. Zhen, C.J. Link Jr., P.M. O’Connor, E. Reed, R. Parker, S.B. Howell, V.A. Bohr, Mol. Cell. Biol. 12 (1992) 3689. [7] Y. Corda, C. Job, M.F. Anin, M. Leng, D. Job, Biochemistry 32 (1993) 8582. [8] E.C.H. Ling, G.W. Allen, T.W. Hambley, J. Am. Chem. Soc. 116 (1994) 2673. [9] T.W. Hambley, E.C.H. Ling, B.A. Messerle, Inorg. Chem. 35 (1996) 4663. [10] T.W. Hambley, E.C.H. Ling, Inorg. Chem., submitted for publication. [11] C.-L. Teo, R.R. Fenton, P. Turner, T.W. Hambley, Aust. J. Chem. (1998) in press. [12] N.F. Curtis, Aust. J. Chem. 39 (1986) 239. [13] J.H. Price, A.N. Williamson, R.F. Schramm, B.B. Wayland, Inorg. Chem. 11 (1972) 1280. [14] teXsan, Crystal Structure Analysis Package, Molecular Structure Corporation, 1985 and 1992. ¨ [15] G.M. Sheldrick, SHELXS-86, in: G.M. Sheldrick, C. Kruger, R. Goddard (Eds.), Crystallographic Computing 3, Oxford University Press, pp. 175–189. [16] D.T. Cromer, J.T. Waber, International Tables for X-ray Crystallography, vol. 4, Kynoch Press, Birmingham, 1974. [17] J.A. Ibers, W.C. Hamilton, Acta Crystallogr. 17 (1964) 781. [18] D.C. Creagh, W.J. McAuley, in: A.J.C. Wilson (Ed.), International Tables for Crystallography, vol. C, Table 4.2.6.8, Kluwer Academic Publishers, Boston, 1992, pp. 219–222. [19] D.C. Creagh, J.H. Hubbell, in: A.J.C. Wilson (Ed.), International Tables for Crystallography, vol. C, Table 4.2.4.3, Kluwer Academic Publishers, Boston, 1992, pp. 200–206. [20] C.K. Johnson, ORTEP, A Thermal Ellipsoid Plotting Program, Oak Ridge National Laboratories, Oak Ridge, 1965. [21] A. Eastman, Biochemistry 21 (1982) 6732. [22] A. Eastman, Pharmac. Ther. 34 (1987) 155. [23] A.M.J. Fichtinger-Schepman, J.L. van der Veer, P.H.M. Lohman, J. Reedijk, J. Inorg. Biochem. 21 (1984) 103. [24] T.W. Hambley, Inorg. Chem. 37 (1998) 3767.