- Email: [email protected]
[1H,15N] NMR studies of the aquation of cis-diamine platinum(II) complexes
Inorganica Chimica Acta 362 (2009) 1022–1026
Contents lists available at ScienceDirect
Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica
Note
[1H,15N] NMR studies of the aquation of cis-diamine platinum(II) complexes Leticia Cubo a, Donald S. Thomas b, Junyong Zhang b, Adoración G. Quiroga a,*, Carmen Navarro-Ranninger a,*, Susan J. Berners-Price b,* a b
Departamento de Química Inorgánica, Universidad Autónoma de Madrid, Madrid, Spain Chemistry M313, School of Biomedical, Biomolecular and Chemical Sciences, The University of Western Australia, 35, Stirling Highway, Crawley, Perth, WA 6009, Australia
a r t i c l e
i n f o
Article history: Received 30 January 2008 Received in revised form 10 March 2008 Accepted 17 March 2008 Available online 8 April 2008 This contribution is dedicated to Professor Bernhard Lippert. Keywords: Platinum drugs [1H,15N] NMR Aquation Kinetics
a b s t r a c t Two 15N-labelled cis-Pt(II) diamine complexes with dimethylamine (15N-dma) and isopropylamine (15Nipa) ligands have been prepared and characterised. [1H,15N] HSQC NMR spectroscopy is used to obtain the rate and equilibrium constants for the aquation of cis-[PtCl2(15N-dma)2] at 298 K in 0.1 M NaClO4 and to determine the pKa values of cis-[PtCl(H2O)(15N-dma)2]+ (6.37) and cis-[Pt(H2O)2(15N-dma)2]2+ (pKa1 = 5.17, pKa2 = 6.47). The rate constants for the first and second aquation steps (k1 = (2.12 ± 0.01) 105 s1, k2 = (8.7 ± 0.7) 106 s1) and anation steps (k1 = (6.7 ± 0.8) 103 M1 s1, k2 = 0.043 ± 0.004 M1 s1) are very similar to those reported for cisplatin under similar conditions, and a minor difference is that slow formation of the hydroxo-bridged dimer is observed. Aquation studies of cis-[PtCl2(15Nipa)2] were precluded by the close proximity of the NH proton signal to the 1H2O resonance. Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved.
1. Introduction An understanding of the aqueous chemistry of diamine Pt(II) anticancer complexes is crucial for establishing their mechanism of action since hydrolysis reactions of chloro Pt(II) complexes are thought to activate them prior to platination of the target site (DNA) and aqua ligands are more labile to substitution than hydroxo ligands [1]. In the continued search for novel platinum anticancer drugs able to circumvent some of the limitations of cisplatin, for example toxicity and resistance, platinum complexes in the trans geometry are of current interest [2]. Since the trans isomer of cisplatin is inactive, it was initially believed that all trans-platinum complexes are ineffective as antitumour agents, but it is now recognised that substitution of NH3 in trans-[PtCl2(L)(L0 )] can give cytotoxicity in the micromolar range, as well as activity in cisplatin resistant cell lines. Included amongst the classes of active transplatinum(II) complexes so far discovered are those with aliphatic amines as spectator ligands [3]. A comparison of the aquation solution chemistry of related cis- and trans-Pt(II) amine complexes is an important first step in understanding their different mechanisms of biological activity. [1H,15N] HSQC NMR spectroscopy has been used extensively to study the hydrolysis of cisplatin and a range of other 15N-labelled * Corresponding authors. Tel.: +61 8 6488 3258;fax: +61 8 6488 1005 (S.J. Berners-Price). E-mail address: [email protected] (S.J. Berners-Price).
platinum am(m)ine complexes [4]. In this report we have used this technique to explore the aquation chemistry of cis-[PtCl2(15Namine)2], for the 15N-labelled amines dimethylamine (dma) and isopropylamine (ipa). The results highlight the influence on the aquation chemistry of classical Pt(II) anticancer complexes in the cis geometry of substituting the NH3 ligands by aliphatic amines. The aquation chemistry of a series of related trans-[PtCl2(15N-amine)(15N-amine0 )] complexes, incorporating 15N-labelled aliphatic amine ligands, will be reported elsewhere [5].
2. Experimental 2.1. Preparation of compounds Starting materials K2PtCl4, methylamine 15N hydrochloride, phthalimide 15N potassium salt, hydrazinium hydroxide and 2bromopropane were purchased from Prolabo, Chambers Hispania, S. L. and Sigma–Aldrich. All solvents chloroform, methanol, tetrahydrofuran (THF), ethyl ether and dimethylformamide (DMF) were purchased from Sigma–Aldrich and purified by standard methods prior to use [6]. 2.1.1. 15N-isopropylamine 15 N-isopropylamine was synthesised via the Gabriel synthesis [7].
0020-1693/$ - see front matter Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2008.03.117
1023
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
2.1.2. Synthesis of cis-[PtCl2(15N-dma)2] and cis-[PtCl2(15N-ipa)2] Full synthetic procedures will be reported elsewhere [5]. Briefly, 15 N-labelled amine (15N-dma or 15N-ipa) was added to a K2PtCl4 suspension in a THF/H2O mixture, using a ratio of 4:1 amine:Pt. After stirring for 14 h in the dark at room temperature, the yellow precipitate was collected, washed with water and chloroform and air dried to give: cis-[PtCl2(15N-dma)2] (56%). Elemental analysis, Anal. Calc. for C4H14N2Cl2Pt: C, 13.48; H, 3.96; N, 7.86. Found: C, 13.66; H, 3.93; N, 7.90%. dH (300.13 MHz; DMSO-d6; Me4Si) 2.47 (6H, d, CH3), 5.64 (1H, d of sept, JHH 5.5, JHN 73.8, NH); dNH (50.68 MHz; 500.13 MHz; DMSO-d6; Me4Si) 29.53, 5.65 (JNPt 368.4, JHPt 69.8); dCH (125.76 MHz; 500.13 MHz; DMSO-d6; Me4Si) 43, 2.6. cis-[PtCl2(15N-ipa)2] (15%). Elemental analysis, Anal. Calc. for C6H18N2Cl2Pt: C, 18.75; H, 4.72; N, 7.29. Found: C, 19.21; H, 4.74; N, 7.30%. dH (300.13 MHz; DMSO-d6; Me4Si) 1.08 (6H, d, JHH 6.3, CH3), 2.97 (1H, sept, JHH 6.45, CH), 4.63 (2H, dd, JHH 6.41, JHN 71.3, NH2); dNH (50.68 MHz; 500.13 MHz; DMSO-d6; Me4Si) 2.89, 4.63 (JNPt 338.5). 2.2. NMR spectroscopy For characterisation of the 15N-labelled complexes NMR spectra were recorded on either Bruker AMX-300 (1H 300.13 MHz) or DRX-500 (1H 500.13 MHz, 15N 50.68 MHz, 13C 125.76 MHz) spectrometers. 1H and 13C chemical shifts are reported relative to tetramethylsilane (Me4Si) or to residual signals of deuterated solvents and 15N NMR spectra are referenced to liquid ammonia [8]. For the aquation studies 1H 1D and 2D [1H,15N] HSQC NMR spectra were recorded at 298 K on a Bruker Avance 600 MHz spectrometer (1H 600.1 MHz, 15N 60.8 MHz) fitted with a pulsed field gradient module and 5 mm triple resonance probehead. 2D [1H,15N] HSQC NMR spectra (optimised for 1J (15N,1H = 72 Hz)) were recorded using the standard Bruker phase sensitive HSQC pulse sequence. 1H chemical shifts were referenced to internal 1,4-dioxane (3.76 ppm), 15N shifts were calibrated externally against 15NH4Cl (1.0 M in 1.0 M HCl in 10% D2O/90% H2O).
Scheme 1.
2.3. pH measurements pH values were determined using a Shindengen pH Boy-P2 (su19A) meter. To avoid leaching of chloride into the bulk sample, aliquots of 5 ll of the solution were placed on the electrode. The meter was calibrated using pH buffers at pH 6.9 and 4.0. Adjustments in pH were made using 0.1 M and 0.01 M HClO4 or 0.1 M and 0.01 M NaOH. 2.4. pKa determination In order to obtain a solution of predominantly monoaqua and diaqua species the following preparative method was performed. AgNO3 (0.68 mg, 4.00 mmol) was dissolved in H2O (1.0 ml) and 1.8 equiv. were added to 1 ml of a solution of cis-[PtCl2(15N-dma)2] (0.75 mg, 2.11 mmol) in 0.113 M NaClO4 in 10% D2O/90% H2O. 1,4dioxane (10 ll) was added and the solution ([Pt] = 2.09 mM) was incubated for 24 h at 298 K and then centrifuged to remove the AgCl precipitate. 1D 1H and 2D [1H,15N] HSQC NMR spectra were recorded in the pH range ca. 3–9.5. Kaleidagraph (Synergy Software, Reading, PA) was used to analyse the pH titration data using the following equations: d ¼ ðdA ½Hþ þ dB K a Þ=ð½Hþ þ K a Þ
ð1Þ
d ¼ ðdAB þ dAA ½Hþ =K a1 þ dBB K a2 =½Hþ Þ=ð1 þ ½Hþ =K a1 þ K a2 =½Hþ Þ ð2Þ where Ka is the acid dissociation constant for the monoaqua species (2), Ka1 and Ka2 those for the diaqua species (3) and dA, dB, dAA, dAB and dBB are the limiting 1H or 15N chemical shifts of
Fig. 1. 2D [1H,15N] HSQC NMR spectrum of cis-[PtCl2(15N-dma)2] at 298 K in 0.1 M NaClO4 (10% D2O/90% H2O) after 69 h at pH 5.4. The peak assignments are shown in Table 1. Note that the 1H and 15N shifts of 2 and 3 shift with pH (Fig. 2).
1024
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
Fig. 2. pH dependence of the 1H and 15N chemical shifts for the (a) monoaqua (open circles, 15N-dma trans to Cl; closed circles, 15N-dma trans to O) and (b) diaqua complexes derived from cis-[PtCl2(15N-dma)2]. The pKa values obtained are (a) d 1H open circles, 6.35 ± 0.02; closed circles, 6.39 ± 0.03, d 15N open circles, 6.29 ± 0.08; closed circles, 6.35 ± 0.04 (b) d 1H pKa1 5.17 ± 0.28; pKa2 6.47 ± 0.10, d 15N pKa1 5.44 ± 0.37; pKa2 6.53 ± 0.20.
cis-[PtCl(H2O)(dma)2]+, cis-[PtCl(OH)(dma)2], cis-[Pt(H2O)2(dcis-[Pt(OH)(H2O)(dma)2]+ and cis-[Pt(OH)2(dma)2], ma)2]2+, respectively. 2.5. Aquation study A fresh solution of cis-[PtCl2(15N-dma)2] (0.44 mg, 1.24 mmol) was prepared in 0.113 M NaClO4 in 10% D2O/90% H2O, and 1,4dioxane (10 ll) added to give a final concentration of 2.04 mM Pt in a total volume of 600 lL. 1D 1H and 2D [1H,15N] HSQC NMR spectra were recorded at 298 K over a period of 69 h. The initial pH was 5.8 and this decreased to 5.4 over the course of the reaction. The kinetic analyses of the aquation reactions were undertaken by measuring the peak volumes in the 2D [1H,15N] HSQC NMR spectra using the Bruker XWIN NMR software and calculating relative concentrations of 15N-species at each time point, as described previously for cisplatin [9]. The aquation data were then analysed using the model shown in Scheme 1. The appropriate differential equations were integrated and rate constants determined by nonlinear optimisation produce using the program SCIENTIST (Version 2.01, MicroMath). The SCIENTIST kinetic model is provided as Supplementary material.
3. Results and discussion 15 N-labelled samples of both cis-[PtCl2(15N-dma)2] and cis[PtCl2(15N-ipa)2] were prepared for hydrolysis studies by [1H,15N] HSQC NMR spectroscopy. For cis-[PtCl2(15N-ipa)2] a single 1H,15N cross-peak at d 4.73/24.7 was observed on dissolution of the complex in water. The 15N shift lies within the expected range
Table 1 1 H and 15N chemical shifts and assignments for cis-[PtCl2(dma)2] and aquation products trans ligand
d 1H
d15N
1 2a
Cl Cl
2b
OH2 OH OH2 OH
5.54 5.78a 5.41a 5.77a 4.96a 6.08b 4.90b 5.40
50.2 47.1a 48.9a 69.2a 60.8a 65.9b 59.2b 62.3
Compound 15
cis-[PtCl2( N-dma)2] cis-[PtCl(OH2)(15N-dma)2]+ cis-[PtCl(OH)(15N-dma)2] cis-[PtCl(OH2)(15N-dma)2]+ cis-[PtCl(OH)(15N-dma)2] cis-[Pt(OH2)2(15N-dma)2]2+ cis-[Pt(OH)2(15N-dma)2] cis-[{Pt(l-OH)(15N-dma)2}2]2+ a,b
3 4
Values for 2 and 3 obtained by fitting Eqs. (1) or (2), respectively, see Fig. 2.
1025
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
Fig. 3. Plots of the relative concentrations of species observed during the aquation of cis-[PtCl2(15N-dma)2] (1) in 0.1 M NaClO4 (pH 5.8) at 298 K. The curves are computer best fits to the rate constants listed in Table 2 (see Scheme 1). Key: 1, open circles; 2, closed triangles; 3, diamonds; 4, closed squares; Cl, open squares.
for 15NH2–Pt–Cl species [4]. The 1H shift lies too close to the 1H2O peak to be able to follow the aquation of this complex and no further studies were undertaken. For cis-[PtCl2(15N-dma)2] the 1H,15N shift is at d 5.54/50.2. While the 15N shift lies well outside the range previously identified for 15NH–Pt–Cl species (ca. d 10) [4] all previous studies have involved the dien ligand in [Pt(15Ndien)Cl] [10] and this appears to be the first example involving a non-chelated NHR2 ligand. The aquation of cis-[PtCl2(15N-dma)2] in 0.1 M NaClO4 at 298 K was followed by [1H,15N] HSQC NMR spectroscopy at pH 5.8. A representative [1H,15N] NMR spectrum (after 69 h) is shown Fig. 1. In the initial [1H,15N] NMR spectrum a single peak for the dichloro complex was observed and after ca. 1 h two new cross-peaks of similar intensity appeared at d 5.70/47.4 and 5.61/67.4 (peaks 2a, 2b, Fig. 1). These peaks can be assigned to the monoaqua com-
plex 2, on the basis of the 15N shifts where a shift change of 20 ppm to higher field is indicative of a chloride ligand being replaced by an oxygen ligand (H2O) trans to an am(m)ine ligand [4]. A single cross-peak at d 5.76/64.3 (3, Fig. 1) appeared after 3.5 h, and is assignable to the diaqua species 3, based on the dependence of the 1H and 15N shifts to pH (Fig. 2). After ca. 7.5 h a new cross-peak appeared at d 5.40/62.3 (4, Fig. 1) which did not shift with pH. The single cross-peak indicates a symmetrical species and similarly in the 1H spectrum only one set of signals is observed for the methyl protons of 4, strongly shielded (by 0.1 ppm) compared to the diaqua species 3 (see Fig. S1). A reasonable assignment is the hydroxo-bridged dimer cis-[{Pt(l-OH)(15N-dma)2}2]2+ as the 15N shift lies in the range between that of trans chloro and aqua ligands, as is found for the hydroxo-bridged dimer of cisplatin [11]. No other peaks appeared until the sample was incubated for a much longer period (9 days). The course of the reaction is slightly different therefore to that found for cisplatin under similar conditions, where two pairs of 1H,15N cross-peaks were assigned to the unsymmetrical hydroxo-bridged dinuclear complexes [{cisPt(15NH3)2}2(l-Cl)(l-OH)]2+ and [{cis-Pt(15NH3)2(OH2)}2(l-OH)]3+, assumed to arise from the monaqua and diaqua complexes, respectively [9]. (see Table 1). The time dependent behaviour of the four species (1–4) is shown in Fig. 3, along with the curves from kinetic fits to the reaction model depicted in Scheme 1 and the rate constants are listed in Table 2. The model assumes that the hydroxo-bridged dimer forms from combination of two diaqua complexes and the good fit obtained further supports the assignment for the hydroxobridged species. Apart from the difference in the minor species, the aquation profile is very similar to that found reported for cisplatin [9] under similar conditions (in 9 mM NaClO4), as reflected by the quite similar values for the equilibrium constants for the first and second aquation steps pK1 and pK2 (Table 2). Indeed, the aquation rate constants for both the first and second steps are remarkably similar for cis-[PtCl2(15N-dma)2] and cisplatin (Table 2), whereas a faster rate of hydrolysis might be expected for a chloride trans to the secondary amine ligand compared to NH3. For example, for the complex cis-[PtCl2NH3(cyclohexylamine)] the rates of aquation are 2–3-fold faster for the chloride ligand trans to the amine ligand (cyclohexamine) compared to trans to NH3, for both the first and second steps [12] and this difference was attributed to the higher trans effect of the amine which has a higher pKa and is a stronger r-donor.
Table 2 Rate and equilibrium constants for the aquation of cis-[PtCl2(15N-dma)2] (Scheme 1) in comparison to cisplatin Compound 15
cis-[PtCl2( N-dma)2] Cisplatina a
k1 (105 s1)
k1 (103 M1 s1)
k2 (106 s1)
k2 (M1 s1)
k3 (M1 s1)
k3 (106 s1)
pK1
pK2
Reference
2.12 ± 0.01 2.38 ± 0.04
6.7 ± 0.8 4.6 ± 0.3
8.7 ± 0.7 14 ± 3
0.043 ± 0.004 0.08 ± 0.02
0.026 ± 0.006
4.8 ± 3.3
2.50 2.07
3.69 3.49
This work [9]
9 mM NaClO4, pH 5.9, 298 K.
Table 3 Comparison of pKa values for some cis-dia(m)mine platinum(II) mono- and di-aqua complexesa Compound
pKa
pKa1
pKa2
Reference
cis-[PtCl2(dimethylamine)2] cis-[PtCl2(NH3)2] cis-[PtCl2NH3(cyclohexylamine)] cis-[PtCl2NH3(2-picoline)] cis-[PtCl2NH3(3-picoline)]
6.37 ± 0.02 6.41 6.73b 6.13/6.49 5.98/6.26
5.17 ± 0.28 5.37 5.68c 5.22 5.07
6.47 ± 0.10 7.21 7.68c 7.16 6.94
This work [15] [17] [18] [18]
a b c
pKa is the value for the monoaqua complexes and pKa1 and pKa2 are the values for the diaqua complexes. Identical values obtained for both monoaqua isomers. Value measured for the aqua ligand trans to NH3.
1026
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
The rate and equilibrium constants for the aquation of a related trans-Pt(II) dimethylamine complex will be reported elsewhere [5] and these data show that the rate of monoaquation of the cis complex is ca. 9-fold slower than the trans complex, whereas the second aquation step is ca. 8-fold faster. This trend is consistent with that of trans-platin where the first aquation step is rapid, whereas the second aquation step is very slow [13]. The faster rate of hydrolysis of trans dichloro complexes compared to the cis isomers has been attributed to the higher trans influence of Cl compared to NH3 [13,14]. [1H,15N] HSQC NMR spectroscopy was also used to determine the pKa values of the aqua ligands in the monoaqua (2) and diaqua (3) complexes. The pH dependences of the 1H and 15N shifts of the NH protons of complexes 2 and 3 over the pH range ca. 3–9.5 are shown in Fig. 2. The 1H and 15N chemical shift changes when deprotonation occurs trans (1H Dd 0.8, 15N Dd 8) or cis (1H Dd 0.3, 15 N Dd 2) to the observed NH group are similar to those observed previously for monoaqua and diaqua forms of cisplatin and other Pt(II) a(m)mine complexes [15–17]. For the monoaqua complex the experimental data were fitted to Eq. (1), to give a pKa value of 6.37 ± 0.02, based on the 1H shifts. These data are more precise than those derived from the 15N shifts due to the poorer resolution in the 15N dimension. The pKa value is very similar to that of the cisplatin monaqua complex (6.41 [15]). It has been noted previously by Sadler [17] that the slighter higher trans influence of an amine ligand compared to ammine might be expected to give a higher pKa for an aqua ligand trans to the amine, but this was not observed for cis-[PtCl(H2O)NH3(cyclohexylamine)] where identical pKa values (6.73 ± 0.02) were found for the two monoaqua isomers [17]. Our results similarly show that the trans influence of a(m)ine ligand appears not to be a major factor in determining the pKa. For the diaqua complex (3) the experimental 1H NMR data were fitted to Eq. (2) to give pKa values of 5.17 ± 0.28 (pKa1) and 6.47 ± 0.10 (pKa2), which were validated by values of 5.44 ± 0.37 and 6.53 ± 0.20, obtained from the 15N data (Fig. 2). As for the monoaqua complex the value of pKa1 is similar to that of the cisplatin diaqua complex (5.37 [15]), but pKa2 is significantly lower (by 0.7 pH units). A difference of ca. 2 log units has been observed for the two pKa values in diaqua complexes of cisplatin and other cis-Pt(II) a(m)mines (see Table 3). The low value for pKa2 could account for the observed formation of the hydroxo-bridged dimer on aquation of cis-[PtCl2(15N-dma)2], but not for cisplatin under similar conditions, with dimer formation favoured by nucleophilic substitution by bound hydroxide.
Acknowledgements This work was supported by the Spanish Ministry of Science and Education (Grants SAF2003-01700 and SAF2006-03296) and the Australian Research Council. L.C. was supported by a FPI doctoral fellowship from Spanish Ministry of Science and Education, Spanish CICYT. Dr. Lindsay Byrne is thanked for assistance with NMR experiments. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2008.03.117. References [1] R.B. Martin, in: B. Lippert (Ed.), Cisplatin: Chemistry and Biochemistry of a Leading Anticancer Drug, Wiley-VCH, Zürich, 1999, p. 183. [2] (a) G. Natile, M. Coluccia, in: A. Sigel, H. Sigel (Eds.), Metal Ions in Biological Systems, Marcel Dekker Inc., New York, 2004, p. 209; (b) G. Natile, M. Coluccia, in: M.J. Clarke, P.J. Sadler (Eds.), Topics in Biological Inorganic Chemistry, Metallopharmaceuticals I, Springer-Verlag, Berlin, 1999, p. 73. [3] E.I. Montero, S. Diaz, A.M. Gonzalez-Vadillo, J.M. Perez, C. Alonso, C. NavarroRanninger, J. Med. Chem. 42 (1999) 4264. [4] S.J. Berners-Price, L. Ronconi, P.J. Sadler, Prog. NMR Spect. 49 (2006) 65. [5] L. Cubo, J. Zhang, A.G. Quiroga, C. Navarro-Ranninger, S.J. Berners-Price, in preparation. [6] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory Chemicals, Butterworth Heinemann, Oxford, 1997. [7] M.S. Gibson, R.W. Bradshaw, Angew. Chem., Int. Ed. 7 (1968) 919. [8] These shifts can be converted to the 15NH4Cl reference used in the [1H,15N] HSQC NMR studies by subtracting 23.6 ppm: D.S. Wishart, C.G. Bigam, J. Yao, F. Abildgaard, H.J. Dyson, E. Oldfield, J.L. Markley, B.D. Sykes, J. Biomol. NMR 6 (1995) 135. [9] M.S. Davies, S.J. Berners-Price, T.W. Hambley, Inorg. Chem. 39 (2000) 5603. [10] Y. Chen, Z. Guo, P.d.S. Murdoch, E. Zang, P.J. Sadler, J. Chem. Soc., Dalton Trans. (1998) 1503. [11] C.J. Boreham, J.A. Broomhead, D.P. Fairlie, Aust. J. Chem. 34 (1981) 659. [12] S.J. Barton, K.J. Barnham, U. Frey, A. Habtemariam, R.E. Sue, P.J. Sadler, Aust. J. Chem. 52 (1999) 173. [13] B. Lippert, Metal Ions Biol. Syst. 33 (1996) 105. [14] G. McGowan, S. Parsons, P.J. Sadler, Inorg. Chem. 44 (2005) 7459. [15] S.J. Berners-Price, T.A. Frenkiel, U. Frey, J.D. Ranford, P.J. Sadler, J. Chem. Soc., Chem. Commun. (1992) 789. [16] M.S. Davies, J.W. Cox, S.J. Berners-Price, W. Barklage, Y. Qu, N. Farrell, Inorg. Chem. 39 (2000) 1710. [17] S.J. Barton, K.J. Barnham, A. Habtemariam, R.E. Sue, P.J. Sadler, Inorg. Chim. Acta 273 (1998) 8. [18] Y. Chen, Z. Guo, S. Parsons, P.J. Sadler, Chem. Eur. J. 4 (1998) 672.
Contents lists available at ScienceDirect
Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica
Note
[1H,15N] NMR studies of the aquation of cis-diamine platinum(II) complexes Leticia Cubo a, Donald S. Thomas b, Junyong Zhang b, Adoración G. Quiroga a,*, Carmen Navarro-Ranninger a,*, Susan J. Berners-Price b,* a b
Departamento de Química Inorgánica, Universidad Autónoma de Madrid, Madrid, Spain Chemistry M313, School of Biomedical, Biomolecular and Chemical Sciences, The University of Western Australia, 35, Stirling Highway, Crawley, Perth, WA 6009, Australia
a r t i c l e
i n f o
Article history: Received 30 January 2008 Received in revised form 10 March 2008 Accepted 17 March 2008 Available online 8 April 2008 This contribution is dedicated to Professor Bernhard Lippert. Keywords: Platinum drugs [1H,15N] NMR Aquation Kinetics
a b s t r a c t Two 15N-labelled cis-Pt(II) diamine complexes with dimethylamine (15N-dma) and isopropylamine (15Nipa) ligands have been prepared and characterised. [1H,15N] HSQC NMR spectroscopy is used to obtain the rate and equilibrium constants for the aquation of cis-[PtCl2(15N-dma)2] at 298 K in 0.1 M NaClO4 and to determine the pKa values of cis-[PtCl(H2O)(15N-dma)2]+ (6.37) and cis-[Pt(H2O)2(15N-dma)2]2+ (pKa1 = 5.17, pKa2 = 6.47). The rate constants for the first and second aquation steps (k1 = (2.12 ± 0.01) 105 s1, k2 = (8.7 ± 0.7) 106 s1) and anation steps (k1 = (6.7 ± 0.8) 103 M1 s1, k2 = 0.043 ± 0.004 M1 s1) are very similar to those reported for cisplatin under similar conditions, and a minor difference is that slow formation of the hydroxo-bridged dimer is observed. Aquation studies of cis-[PtCl2(15Nipa)2] were precluded by the close proximity of the NH proton signal to the 1H2O resonance. Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved.
1. Introduction An understanding of the aqueous chemistry of diamine Pt(II) anticancer complexes is crucial for establishing their mechanism of action since hydrolysis reactions of chloro Pt(II) complexes are thought to activate them prior to platination of the target site (DNA) and aqua ligands are more labile to substitution than hydroxo ligands [1]. In the continued search for novel platinum anticancer drugs able to circumvent some of the limitations of cisplatin, for example toxicity and resistance, platinum complexes in the trans geometry are of current interest [2]. Since the trans isomer of cisplatin is inactive, it was initially believed that all trans-platinum complexes are ineffective as antitumour agents, but it is now recognised that substitution of NH3 in trans-[PtCl2(L)(L0 )] can give cytotoxicity in the micromolar range, as well as activity in cisplatin resistant cell lines. Included amongst the classes of active transplatinum(II) complexes so far discovered are those with aliphatic amines as spectator ligands [3]. A comparison of the aquation solution chemistry of related cis- and trans-Pt(II) amine complexes is an important first step in understanding their different mechanisms of biological activity. [1H,15N] HSQC NMR spectroscopy has been used extensively to study the hydrolysis of cisplatin and a range of other 15N-labelled * Corresponding authors. Tel.: +61 8 6488 3258;fax: +61 8 6488 1005 (S.J. Berners-Price). E-mail address: [email protected] (S.J. Berners-Price).
platinum am(m)ine complexes [4]. In this report we have used this technique to explore the aquation chemistry of cis-[PtCl2(15Namine)2], for the 15N-labelled amines dimethylamine (dma) and isopropylamine (ipa). The results highlight the influence on the aquation chemistry of classical Pt(II) anticancer complexes in the cis geometry of substituting the NH3 ligands by aliphatic amines. The aquation chemistry of a series of related trans-[PtCl2(15N-amine)(15N-amine0 )] complexes, incorporating 15N-labelled aliphatic amine ligands, will be reported elsewhere [5].
2. Experimental 2.1. Preparation of compounds Starting materials K2PtCl4, methylamine 15N hydrochloride, phthalimide 15N potassium salt, hydrazinium hydroxide and 2bromopropane were purchased from Prolabo, Chambers Hispania, S. L. and Sigma–Aldrich. All solvents chloroform, methanol, tetrahydrofuran (THF), ethyl ether and dimethylformamide (DMF) were purchased from Sigma–Aldrich and purified by standard methods prior to use [6]. 2.1.1. 15N-isopropylamine 15 N-isopropylamine was synthesised via the Gabriel synthesis [7].
0020-1693/$ - see front matter Crown Copyright Ó 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2008.03.117
1023
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
2.1.2. Synthesis of cis-[PtCl2(15N-dma)2] and cis-[PtCl2(15N-ipa)2] Full synthetic procedures will be reported elsewhere [5]. Briefly, 15 N-labelled amine (15N-dma or 15N-ipa) was added to a K2PtCl4 suspension in a THF/H2O mixture, using a ratio of 4:1 amine:Pt. After stirring for 14 h in the dark at room temperature, the yellow precipitate was collected, washed with water and chloroform and air dried to give: cis-[PtCl2(15N-dma)2] (56%). Elemental analysis, Anal. Calc. for C4H14N2Cl2Pt: C, 13.48; H, 3.96; N, 7.86. Found: C, 13.66; H, 3.93; N, 7.90%. dH (300.13 MHz; DMSO-d6; Me4Si) 2.47 (6H, d, CH3), 5.64 (1H, d of sept, JHH 5.5, JHN 73.8, NH); dNH (50.68 MHz; 500.13 MHz; DMSO-d6; Me4Si) 29.53, 5.65 (JNPt 368.4, JHPt 69.8); dCH (125.76 MHz; 500.13 MHz; DMSO-d6; Me4Si) 43, 2.6. cis-[PtCl2(15N-ipa)2] (15%). Elemental analysis, Anal. Calc. for C6H18N2Cl2Pt: C, 18.75; H, 4.72; N, 7.29. Found: C, 19.21; H, 4.74; N, 7.30%. dH (300.13 MHz; DMSO-d6; Me4Si) 1.08 (6H, d, JHH 6.3, CH3), 2.97 (1H, sept, JHH 6.45, CH), 4.63 (2H, dd, JHH 6.41, JHN 71.3, NH2); dNH (50.68 MHz; 500.13 MHz; DMSO-d6; Me4Si) 2.89, 4.63 (JNPt 338.5). 2.2. NMR spectroscopy For characterisation of the 15N-labelled complexes NMR spectra were recorded on either Bruker AMX-300 (1H 300.13 MHz) or DRX-500 (1H 500.13 MHz, 15N 50.68 MHz, 13C 125.76 MHz) spectrometers. 1H and 13C chemical shifts are reported relative to tetramethylsilane (Me4Si) or to residual signals of deuterated solvents and 15N NMR spectra are referenced to liquid ammonia [8]. For the aquation studies 1H 1D and 2D [1H,15N] HSQC NMR spectra were recorded at 298 K on a Bruker Avance 600 MHz spectrometer (1H 600.1 MHz, 15N 60.8 MHz) fitted with a pulsed field gradient module and 5 mm triple resonance probehead. 2D [1H,15N] HSQC NMR spectra (optimised for 1J (15N,1H = 72 Hz)) were recorded using the standard Bruker phase sensitive HSQC pulse sequence. 1H chemical shifts were referenced to internal 1,4-dioxane (3.76 ppm), 15N shifts were calibrated externally against 15NH4Cl (1.0 M in 1.0 M HCl in 10% D2O/90% H2O).
Scheme 1.
2.3. pH measurements pH values were determined using a Shindengen pH Boy-P2 (su19A) meter. To avoid leaching of chloride into the bulk sample, aliquots of 5 ll of the solution were placed on the electrode. The meter was calibrated using pH buffers at pH 6.9 and 4.0. Adjustments in pH were made using 0.1 M and 0.01 M HClO4 or 0.1 M and 0.01 M NaOH. 2.4. pKa determination In order to obtain a solution of predominantly monoaqua and diaqua species the following preparative method was performed. AgNO3 (0.68 mg, 4.00 mmol) was dissolved in H2O (1.0 ml) and 1.8 equiv. were added to 1 ml of a solution of cis-[PtCl2(15N-dma)2] (0.75 mg, 2.11 mmol) in 0.113 M NaClO4 in 10% D2O/90% H2O. 1,4dioxane (10 ll) was added and the solution ([Pt] = 2.09 mM) was incubated for 24 h at 298 K and then centrifuged to remove the AgCl precipitate. 1D 1H and 2D [1H,15N] HSQC NMR spectra were recorded in the pH range ca. 3–9.5. Kaleidagraph (Synergy Software, Reading, PA) was used to analyse the pH titration data using the following equations: d ¼ ðdA ½Hþ þ dB K a Þ=ð½Hþ þ K a Þ
ð1Þ
d ¼ ðdAB þ dAA ½Hþ =K a1 þ dBB K a2 =½Hþ Þ=ð1 þ ½Hþ =K a1 þ K a2 =½Hþ Þ ð2Þ where Ka is the acid dissociation constant for the monoaqua species (2), Ka1 and Ka2 those for the diaqua species (3) and dA, dB, dAA, dAB and dBB are the limiting 1H or 15N chemical shifts of
Fig. 1. 2D [1H,15N] HSQC NMR spectrum of cis-[PtCl2(15N-dma)2] at 298 K in 0.1 M NaClO4 (10% D2O/90% H2O) after 69 h at pH 5.4. The peak assignments are shown in Table 1. Note that the 1H and 15N shifts of 2 and 3 shift with pH (Fig. 2).
1024
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
Fig. 2. pH dependence of the 1H and 15N chemical shifts for the (a) monoaqua (open circles, 15N-dma trans to Cl; closed circles, 15N-dma trans to O) and (b) diaqua complexes derived from cis-[PtCl2(15N-dma)2]. The pKa values obtained are (a) d 1H open circles, 6.35 ± 0.02; closed circles, 6.39 ± 0.03, d 15N open circles, 6.29 ± 0.08; closed circles, 6.35 ± 0.04 (b) d 1H pKa1 5.17 ± 0.28; pKa2 6.47 ± 0.10, d 15N pKa1 5.44 ± 0.37; pKa2 6.53 ± 0.20.
cis-[PtCl(H2O)(dma)2]+, cis-[PtCl(OH)(dma)2], cis-[Pt(H2O)2(dcis-[Pt(OH)(H2O)(dma)2]+ and cis-[Pt(OH)2(dma)2], ma)2]2+, respectively. 2.5. Aquation study A fresh solution of cis-[PtCl2(15N-dma)2] (0.44 mg, 1.24 mmol) was prepared in 0.113 M NaClO4 in 10% D2O/90% H2O, and 1,4dioxane (10 ll) added to give a final concentration of 2.04 mM Pt in a total volume of 600 lL. 1D 1H and 2D [1H,15N] HSQC NMR spectra were recorded at 298 K over a period of 69 h. The initial pH was 5.8 and this decreased to 5.4 over the course of the reaction. The kinetic analyses of the aquation reactions were undertaken by measuring the peak volumes in the 2D [1H,15N] HSQC NMR spectra using the Bruker XWIN NMR software and calculating relative concentrations of 15N-species at each time point, as described previously for cisplatin [9]. The aquation data were then analysed using the model shown in Scheme 1. The appropriate differential equations were integrated and rate constants determined by nonlinear optimisation produce using the program SCIENTIST (Version 2.01, MicroMath). The SCIENTIST kinetic model is provided as Supplementary material.
3. Results and discussion 15 N-labelled samples of both cis-[PtCl2(15N-dma)2] and cis[PtCl2(15N-ipa)2] were prepared for hydrolysis studies by [1H,15N] HSQC NMR spectroscopy. For cis-[PtCl2(15N-ipa)2] a single 1H,15N cross-peak at d 4.73/24.7 was observed on dissolution of the complex in water. The 15N shift lies within the expected range
Table 1 1 H and 15N chemical shifts and assignments for cis-[PtCl2(dma)2] and aquation products trans ligand
d 1H
d15N
1 2a
Cl Cl
2b
OH2 OH OH2 OH
5.54 5.78a 5.41a 5.77a 4.96a 6.08b 4.90b 5.40
50.2 47.1a 48.9a 69.2a 60.8a 65.9b 59.2b 62.3
Compound 15
cis-[PtCl2( N-dma)2] cis-[PtCl(OH2)(15N-dma)2]+ cis-[PtCl(OH)(15N-dma)2] cis-[PtCl(OH2)(15N-dma)2]+ cis-[PtCl(OH)(15N-dma)2] cis-[Pt(OH2)2(15N-dma)2]2+ cis-[Pt(OH)2(15N-dma)2] cis-[{Pt(l-OH)(15N-dma)2}2]2+ a,b
3 4
Values for 2 and 3 obtained by fitting Eqs. (1) or (2), respectively, see Fig. 2.
1025
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
Fig. 3. Plots of the relative concentrations of species observed during the aquation of cis-[PtCl2(15N-dma)2] (1) in 0.1 M NaClO4 (pH 5.8) at 298 K. The curves are computer best fits to the rate constants listed in Table 2 (see Scheme 1). Key: 1, open circles; 2, closed triangles; 3, diamonds; 4, closed squares; Cl, open squares.
for 15NH2–Pt–Cl species [4]. The 1H shift lies too close to the 1H2O peak to be able to follow the aquation of this complex and no further studies were undertaken. For cis-[PtCl2(15N-dma)2] the 1H,15N shift is at d 5.54/50.2. While the 15N shift lies well outside the range previously identified for 15NH–Pt–Cl species (ca. d 10) [4] all previous studies have involved the dien ligand in [Pt(15Ndien)Cl] [10] and this appears to be the first example involving a non-chelated NHR2 ligand. The aquation of cis-[PtCl2(15N-dma)2] in 0.1 M NaClO4 at 298 K was followed by [1H,15N] HSQC NMR spectroscopy at pH 5.8. A representative [1H,15N] NMR spectrum (after 69 h) is shown Fig. 1. In the initial [1H,15N] NMR spectrum a single peak for the dichloro complex was observed and after ca. 1 h two new cross-peaks of similar intensity appeared at d 5.70/47.4 and 5.61/67.4 (peaks 2a, 2b, Fig. 1). These peaks can be assigned to the monoaqua com-
plex 2, on the basis of the 15N shifts where a shift change of 20 ppm to higher field is indicative of a chloride ligand being replaced by an oxygen ligand (H2O) trans to an am(m)ine ligand [4]. A single cross-peak at d 5.76/64.3 (3, Fig. 1) appeared after 3.5 h, and is assignable to the diaqua species 3, based on the dependence of the 1H and 15N shifts to pH (Fig. 2). After ca. 7.5 h a new cross-peak appeared at d 5.40/62.3 (4, Fig. 1) which did not shift with pH. The single cross-peak indicates a symmetrical species and similarly in the 1H spectrum only one set of signals is observed for the methyl protons of 4, strongly shielded (by 0.1 ppm) compared to the diaqua species 3 (see Fig. S1). A reasonable assignment is the hydroxo-bridged dimer cis-[{Pt(l-OH)(15N-dma)2}2]2+ as the 15N shift lies in the range between that of trans chloro and aqua ligands, as is found for the hydroxo-bridged dimer of cisplatin [11]. No other peaks appeared until the sample was incubated for a much longer period (9 days). The course of the reaction is slightly different therefore to that found for cisplatin under similar conditions, where two pairs of 1H,15N cross-peaks were assigned to the unsymmetrical hydroxo-bridged dinuclear complexes [{cisPt(15NH3)2}2(l-Cl)(l-OH)]2+ and [{cis-Pt(15NH3)2(OH2)}2(l-OH)]3+, assumed to arise from the monaqua and diaqua complexes, respectively [9]. (see Table 1). The time dependent behaviour of the four species (1–4) is shown in Fig. 3, along with the curves from kinetic fits to the reaction model depicted in Scheme 1 and the rate constants are listed in Table 2. The model assumes that the hydroxo-bridged dimer forms from combination of two diaqua complexes and the good fit obtained further supports the assignment for the hydroxobridged species. Apart from the difference in the minor species, the aquation profile is very similar to that found reported for cisplatin [9] under similar conditions (in 9 mM NaClO4), as reflected by the quite similar values for the equilibrium constants for the first and second aquation steps pK1 and pK2 (Table 2). Indeed, the aquation rate constants for both the first and second steps are remarkably similar for cis-[PtCl2(15N-dma)2] and cisplatin (Table 2), whereas a faster rate of hydrolysis might be expected for a chloride trans to the secondary amine ligand compared to NH3. For example, for the complex cis-[PtCl2NH3(cyclohexylamine)] the rates of aquation are 2–3-fold faster for the chloride ligand trans to the amine ligand (cyclohexamine) compared to trans to NH3, for both the first and second steps [12] and this difference was attributed to the higher trans effect of the amine which has a higher pKa and is a stronger r-donor.
Table 2 Rate and equilibrium constants for the aquation of cis-[PtCl2(15N-dma)2] (Scheme 1) in comparison to cisplatin Compound 15
cis-[PtCl2( N-dma)2] Cisplatina a
k1 (105 s1)
k1 (103 M1 s1)
k2 (106 s1)
k2 (M1 s1)
k3 (M1 s1)
k3 (106 s1)
pK1
pK2
Reference
2.12 ± 0.01 2.38 ± 0.04
6.7 ± 0.8 4.6 ± 0.3
8.7 ± 0.7 14 ± 3
0.043 ± 0.004 0.08 ± 0.02
0.026 ± 0.006
4.8 ± 3.3
2.50 2.07
3.69 3.49
This work [9]
9 mM NaClO4, pH 5.9, 298 K.
Table 3 Comparison of pKa values for some cis-dia(m)mine platinum(II) mono- and di-aqua complexesa Compound
pKa
pKa1
pKa2
Reference
cis-[PtCl2(dimethylamine)2] cis-[PtCl2(NH3)2] cis-[PtCl2NH3(cyclohexylamine)] cis-[PtCl2NH3(2-picoline)] cis-[PtCl2NH3(3-picoline)]
6.37 ± 0.02 6.41 6.73b 6.13/6.49 5.98/6.26
5.17 ± 0.28 5.37 5.68c 5.22 5.07
6.47 ± 0.10 7.21 7.68c 7.16 6.94
This work [15] [17] [18] [18]
a b c
pKa is the value for the monoaqua complexes and pKa1 and pKa2 are the values for the diaqua complexes. Identical values obtained for both monoaqua isomers. Value measured for the aqua ligand trans to NH3.
1026
L. Cubo et al. / Inorganica Chimica Acta 362 (2009) 1022–1026
The rate and equilibrium constants for the aquation of a related trans-Pt(II) dimethylamine complex will be reported elsewhere [5] and these data show that the rate of monoaquation of the cis complex is ca. 9-fold slower than the trans complex, whereas the second aquation step is ca. 8-fold faster. This trend is consistent with that of trans-platin where the first aquation step is rapid, whereas the second aquation step is very slow [13]. The faster rate of hydrolysis of trans dichloro complexes compared to the cis isomers has been attributed to the higher trans influence of Cl compared to NH3 [13,14]. [1H,15N] HSQC NMR spectroscopy was also used to determine the pKa values of the aqua ligands in the monoaqua (2) and diaqua (3) complexes. The pH dependences of the 1H and 15N shifts of the NH protons of complexes 2 and 3 over the pH range ca. 3–9.5 are shown in Fig. 2. The 1H and 15N chemical shift changes when deprotonation occurs trans (1H Dd 0.8, 15N Dd 8) or cis (1H Dd 0.3, 15 N Dd 2) to the observed NH group are similar to those observed previously for monoaqua and diaqua forms of cisplatin and other Pt(II) a(m)mine complexes [15–17]. For the monoaqua complex the experimental data were fitted to Eq. (1), to give a pKa value of 6.37 ± 0.02, based on the 1H shifts. These data are more precise than those derived from the 15N shifts due to the poorer resolution in the 15N dimension. The pKa value is very similar to that of the cisplatin monaqua complex (6.41 [15]). It has been noted previously by Sadler [17] that the slighter higher trans influence of an amine ligand compared to ammine might be expected to give a higher pKa for an aqua ligand trans to the amine, but this was not observed for cis-[PtCl(H2O)NH3(cyclohexylamine)] where identical pKa values (6.73 ± 0.02) were found for the two monoaqua isomers [17]. Our results similarly show that the trans influence of a(m)ine ligand appears not to be a major factor in determining the pKa. For the diaqua complex (3) the experimental 1H NMR data were fitted to Eq. (2) to give pKa values of 5.17 ± 0.28 (pKa1) and 6.47 ± 0.10 (pKa2), which were validated by values of 5.44 ± 0.37 and 6.53 ± 0.20, obtained from the 15N data (Fig. 2). As for the monoaqua complex the value of pKa1 is similar to that of the cisplatin diaqua complex (5.37 [15]), but pKa2 is significantly lower (by 0.7 pH units). A difference of ca. 2 log units has been observed for the two pKa values in diaqua complexes of cisplatin and other cis-Pt(II) a(m)mines (see Table 3). The low value for pKa2 could account for the observed formation of the hydroxo-bridged dimer on aquation of cis-[PtCl2(15N-dma)2], but not for cisplatin under similar conditions, with dimer formation favoured by nucleophilic substitution by bound hydroxide.
Acknowledgements This work was supported by the Spanish Ministry of Science and Education (Grants SAF2003-01700 and SAF2006-03296) and the Australian Research Council. L.C. was supported by a FPI doctoral fellowship from Spanish Ministry of Science and Education, Spanish CICYT. Dr. Lindsay Byrne is thanked for assistance with NMR experiments. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2008.03.117. References [1] R.B. Martin, in: B. Lippert (Ed.), Cisplatin: Chemistry and Biochemistry of a Leading Anticancer Drug, Wiley-VCH, Zürich, 1999, p. 183. [2] (a) G. Natile, M. Coluccia, in: A. Sigel, H. Sigel (Eds.), Metal Ions in Biological Systems, Marcel Dekker Inc., New York, 2004, p. 209; (b) G. Natile, M. Coluccia, in: M.J. Clarke, P.J. Sadler (Eds.), Topics in Biological Inorganic Chemistry, Metallopharmaceuticals I, Springer-Verlag, Berlin, 1999, p. 73. [3] E.I. Montero, S. Diaz, A.M. Gonzalez-Vadillo, J.M. Perez, C. Alonso, C. NavarroRanninger, J. Med. Chem. 42 (1999) 4264. [4] S.J. Berners-Price, L. Ronconi, P.J. Sadler, Prog. NMR Spect. 49 (2006) 65. [5] L. Cubo, J. Zhang, A.G. Quiroga, C. Navarro-Ranninger, S.J. Berners-Price, in preparation. [6] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory Chemicals, Butterworth Heinemann, Oxford, 1997. [7] M.S. Gibson, R.W. Bradshaw, Angew. Chem., Int. Ed. 7 (1968) 919. [8] These shifts can be converted to the 15NH4Cl reference used in the [1H,15N] HSQC NMR studies by subtracting 23.6 ppm: D.S. Wishart, C.G. Bigam, J. Yao, F. Abildgaard, H.J. Dyson, E. Oldfield, J.L. Markley, B.D. Sykes, J. Biomol. NMR 6 (1995) 135. [9] M.S. Davies, S.J. Berners-Price, T.W. Hambley, Inorg. Chem. 39 (2000) 5603. [10] Y. Chen, Z. Guo, P.d.S. Murdoch, E. Zang, P.J. Sadler, J. Chem. Soc., Dalton Trans. (1998) 1503. [11] C.J. Boreham, J.A. Broomhead, D.P. Fairlie, Aust. J. Chem. 34 (1981) 659. [12] S.J. Barton, K.J. Barnham, U. Frey, A. Habtemariam, R.E. Sue, P.J. Sadler, Aust. J. Chem. 52 (1999) 173. [13] B. Lippert, Metal Ions Biol. Syst. 33 (1996) 105. [14] G. McGowan, S. Parsons, P.J. Sadler, Inorg. Chem. 44 (2005) 7459. [15] S.J. Berners-Price, T.A. Frenkiel, U. Frey, J.D. Ranford, P.J. Sadler, J. Chem. Soc., Chem. Commun. (1992) 789. [16] M.S. Davies, J.W. Cox, S.J. Berners-Price, W. Barklage, Y. Qu, N. Farrell, Inorg. Chem. 39 (2000) 1710. [17] S.J. Barton, K.J. Barnham, A. Habtemariam, R.E. Sue, P.J. Sadler, Inorg. Chim. Acta 273 (1998) 8. [18] Y. Chen, Z. Guo, S. Parsons, P.J. Sadler, Chem. Eur. J. 4 (1998) 672.