Kinetics and mechanism for reduction of oral anticancer platinum(iv) dicarboxylate compounds by L-ascorbate ions

Article abstract of DOI:10.1039/a909484i

Ascorbate(Asc) reductions of the oral anticancer platinum(iv) prodrugs m,trans,cu-[PtCl₂(OAc)₂(cha)(NH ₃)](JM216) and cis, trans;,cis-[PtCl₂(OCOC₃H₇) ₂(cha)(NH₃)](JM221) and of the isomers of JM216, viz.trans,cis,cis-[PtCl2(OAc)₂(cha)(NH₃)](JM394) and trans,trans,trans-[PtCI₂(OAc)₂(cha)(NH ₃)](JM576)(OAc = acetate, cha = cyclohexylamine) have been investigated in a 1.0 M aqueous perchlorate medium using stopped-flow and conventional UV/VIS spectrophotometry as a function of temperature and pH. JM216 and 221 are reduced to ra-[PtCl₂(cha)(NH₃)](JM118) and JM394 and 576 to CM- and trans-[Pt(OAc)₂(cha)(NH₃)], respectively. The redox reactions follow the second-order rate law: -d[Pt(iv)]/dt = k [Pt(iv)] [Asc]_tot where k is a pH dependent second-order overall rate constant and [Asc]_tot = [Asc^2-] + [H₂Asc] + [H₂Asc]. Reduction of JM216 and JM221 is slow(overall rate constants k²⁹⁸ = 5.08 ±10^-2 and 3.25 × 1(T2 mol^-1 dm³ s^-1 at pH 7.12, respectively) and is suggested to take place via an outer-sphere mechanism. Reductions of JM394 and JM576 are more than three orders of magnitude faster(k²⁹⁸ = 230 ±6 mol^-1 dm3 s^-1 at pH 7.0 for JM394). They are suggested to take place by a mechanism involving a reductive attack on one of the mutually trans chloride ligands by Asc² and less efficiently by H Asc^- leading to the formation of a chloride-bridged activated complex. The second-order rate constants for reduction of JM394 by HAsc^- and Asc^2- at 25 °C are 0.548 ±0.004 and(4.46 ±0.01) × 10⁶ mol^-1 dm3 s^-1, respectively. The rate constants for reduction of JM216 and JM221 by Asc^2- at 25 °C are calculated to be 672 ±15 and 428 ±10 mol1 dm3 s^-1, respectively and reduction by HAsc^- was not observed under these conditions. Thus, Asc^2- is up to 7 orders of magnitude more efficient as a reductant than HAsc^-. H2Asc is virtually inactive. The activation parameters Asc^2- and AS⁺ for reduction of JM216, JM221, JM394, and JM576 by Asc^2- are 52 ±1,46 ±1, 56.2 ±0.5, and 63 ±2 kJ mol^-1 and -97 ±4, -120 ±4, -24 ±2, and -8 ±5 J K^-1 mol^-1, respectively. An isokinetic relationship gives further support to the mechanistic assignments. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/a909484i

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetics and mechanism for reduction of oral anticancer
platinum(IV) dicarboxylate compounds by L-ascorbate ions†
Kelemu Lemma,a,c Alan M. Sargesonb and Lars I. Elding*a
a Inorganic Chemistry 1, Chemical Center, Lund University, PO Box 124, SE-221 00 Lund,
Sweden. E-mail: LarsI.Elding@inorg.lu.se
b Research School of Chemistry, A. N. U., Canberra ACT 0200, Australia
c Department of Chemistry, Addis Ababa University, PO Box 1176, Addis Ababa, Ethiopia
Received 1st December 1999, Accepted 11th February 2000
Published on the Web 10th March 2000
Ascorbate (Asc) reductions of the oral anticancer platinum() prodrugs cis,trans,cis-[PtCl2(OAc)2(cha)(NH3)]
(JM216) and cis,trans,cis-[PtCl2(OCOC3H7)2(cha)(NH3)] (JM221) and of the isomers of JM216, viz. trans,cis,cis-
[PtCl2(OAc)2(cha)(NH3)] (JM394) and trans,trans,trans-[PtCl2(OAc)2(cha)(NH3)] (JM576) (OAc = acetate,
cha = cyclohexylamine) have been investigated in a 1.0 M aqueous perchlorate medium using stopped-flow and
conventional UV/VIS spectrophotometry as a function of temperature and pH. JM216 and 221 are reduced to
cis-[PtCl2(cha)(NH3)] (JM118) and JM394 and 576 to cis- and trans-[Pt(OAc)2(cha)(NH3)], respectively. The redox
reactions follow the second-order rate law: Ϫd[Pt()]/dt = k [Pt()] [Asc]tot where k is a pH dependent second-order
overall rate constant and [Asc]tot = [Asc2Ϫ] ϩ [HAscϪ] ϩ [H2Asc]. Reduction of JM216 and JM221 is slow (overall rate
constants k298 = 5.08 10Ϫ2 and 3.25 × 10Ϫ2 molϪ1 dm3 sϪ1 at pH 7.12, respectively) and is suggested to take place
via an outer-sphere mechanism. Reductions of JM394 and JM576 are more than three orders of magnitude faster
(k298 = 230 6 molϪ1 dm3 sϪ1 at pH 7.0 for JM394). They are suggested to take place by a mechanism involving a
reductive attack on one of the mutually trans chloride ligands by Asc2Ϫ and less efficiently by HAscϪ leading to the
formation of a chloride-bridged activated complex. The second-order rate constants for reduction of JM394 by
HAscϪ and Asc2Ϫ at 25 ЊC are 0.548 0.004 and (4.46 0.01) × 106 molϪ1 dm3 sϪ1, respectively. The rate constants
for reduction of JM216 and JM221 by Asc2Ϫ at 25 ЊC are calculated to be 672 15 and 428 10 molϪ1 dm3 sϪ1
,
respectively and reduction by HAscϪ was not observed under these conditions. Thus, Asc2Ϫ is up to 7 orders of
magnitude more efficient as a reductant than HAscϪ. H2Asc is virtually inactive. The activation parameters ∆H‡ and
∆S‡ for reduction of JM216, JM221, JM394, and JM576 by Asc2Ϫ are 52 1, 46 1, 56.2 0.5, and 63 2 kJ molϪ1
and Ϫ97 4, Ϫ120 4, Ϫ24 2, and Ϫ8 5 J KϪ1 molϪ1, respectively. An isokinetic relationship gives further
support to the mechanistic assignments.
compounds in cisplatin-resistant cells may be a consequence of
Introduction
their lipophilic nature facilitating transport into the cell.4,6
The current search for platinum-based anticancer active com-
It is almost certain that cisplatin exerts its cytotoxicity
pounds aims at developing complexes which do not show
through binding to DNA, producing both intra- and inter-
cross-resistance with the widely used drugs cis-[Pt(NH3)2Cl2]
strand cross-links which inhibit replication.7 Since platinum()
(cisplatin) and cis-diammine(1,1-cyclobutanedicarboxylato)-
complexes are generally substitution inert,8 it is presumed that
platinum() (carboplatin), which are active against a broader
they are reduced to the more reactive platinum() analogues
range of tumours, and which can be taken orally. Oral drugs are
before interaction with DNA. Numerous experiments have
simpler to use and less expensive than those requiring intra-
venous administration. A novel class of drugs, ammine/
amine(dichloro)platinum() dicarboxylates, with suitable
shown that anticancer active and other platinum() complexes
are reduced by both extra- and intra-cellular reducing agents.9–12
JM216,forinstance,isreportedtoundergobio-transformationin
properties for oral administration and antitumour activity in
the plasma into at least six different metabolites amongst which
cisplatin-resistant human cancer cells has been developed.1,2
cis-[PtCl2(cha)(NH3)] (JM118) is a major product.13 Similarly,
Examples include cis,trans,cis-[PtCl2(OAc)2(cha)(NH3)] (JM-
JM221 displays a marked time dependent cytotoxicity in
216)
(cha = cyclohexylamine)
and
cis,trans,cis-[PtCl2-
human ovarian carcinoma cells, implying that a slow reductive
and hydrolytic activation of the drug might be required for
activity.14
(OCOC3H7)2(cha)(NH3)] (JM221) which are lipophilic and
robust enough to survive the gastric environment.3 In par-
ticular, they are effective in cells where resistance to other
platinum drugs is due to a decreased uptake of platinum.4 The
precise mechanism whereby cisplatin enters the cells is not clear.
Passive diffusion as well as carrier mediated mechanisms have
been proposed.5 The effectiveness of the Pt() dicarboxylate
The anticancer activity of the platinum() dicarboxylate
compounds is likely to be due to effective platinum transport
into the cell followed by reduction to the more reactive plat-
inum() compounds. Knowledge of the reactivity of the
platinum() compounds towards reduction by potential
bio-reductants such as ascorbic acid and glutathione may be
important for the understanding of the mechanism of their
antitumour activity as well as for the design of new compounds
with suitable pharmacokinetic properties.15 Choi and co-
workers recently studied ascorbate reduction of a series of
† Electronic supplementary information (ESI) available: 1H NMR spec-
tra of JM576, JM216 and their reduction products with ascorbic acid
and 1H NMR spectra of 2 mmol dmϪ3 JM576 with 200 mmol dmϪ3
chloride recorded after 1.5 and 72 h. See http://www.rsc.org/suppdata/
dt/a9/a909484i/
DOI: 10.1039/a909484i
J. Chem. Soc., Dalton Trans., 2000, 1167–1172
This journal is © The Royal Society of Chemistry 2000
1167

View Article Online
Pt() anticancer active complexes, investigating the effects of
axial and carrier ligands on the relative reduction and cyto-
toxicity,16 and Ranford and co-workers have investigated the
reduction of a bis(carboxylato) complex by methionine and
cysteine.17
Physical measurements
The pH of the buffer solutions was measured with a Metrohm
632 digital pH meter equipped with a combination Metrohm
glass electrode and activities of oxonium ions aH = 10ϪpH were
calculated from direct pH meter readings. UV/VIS spectra were
recorded with a CARY 300 Bio UV/VIS spectrophotometer
using 1.0 cm Quartz Suprasil cells. Proton NMR spectra were
recorded with a Varian Unity 300 spectrometer operating at
a frequency of 299.779 MHz with D2O as solvent and the
residual solvent signal as a reference at constant pH, ionic
medium and temperature.
The study by Choi et al.16 did not focus on elucidating the
details of the molecular redox mechanism (vide infra). In view
of this limitation and the significance of JM216 and JM221 as
oral anticancer prodrugs, it was considered worthwhile to under-
take a kinetic and mechanistic investigation on the reduction of
JM216 and JM221 by ascorbic acid. In order to obtain further
insight into the mechanism, the reduction of trans,cis,cis-
[PtCl2(OAc)2(cha)(NH3)] (JM394) and trans,trans,trans-[PtCl2-
(OAc)2(cha)(NH3)] (JM576) have also been studied. The
structures of the platinum() compounds and -ascorbic acid
are shown in Chart 1. Attempts to study the reduction of
Kinetic measurements
The redox reactions were investigated in a 1.0 mol dmϪ3 aque-
ous perchlorate medium using stopped-flow and conventional
UV/VIS spectrophotometry. All kinetic measurements were
performed under pseudo-first-order conditions with at least a
10-fold excess of ascorbic acid by monitoring the absorbance
decrease at 310 (JM394 and JM576) and 318 nm (JM216 and
JM221). Three to six repetitive measurements were made and
reactions were followed for at least 4 half-lives. The reduc-
tion of JM394 and JM576 was investigated in the region
4.0 ≤ pH ≤ 7.0 at 25 ЊC in acetate or phosphate buffers using an
Applied Photophysics SX-18MV Stopped-Flow ASVD spectro-
photometer, and that of JM216 and JM221 at 35 ЊC with
7.0 ≤ pH ≤ 7.5 in TRIS buffer using the Cary 300 Bio UV/VIS
spectrophotometer. For JM-216 and -221, 1.00 cm3 of metal
solution in water was mixed with 1.00 cm3 of NaHAsc dis-
solved in TRIS buffer of total ionic strength 2.0 mol dmϪ3
(NaClO4) directly in a 1.00 cm spectrophotometer cell. No
reactions between the buffers and the platinum complexes were
observed during the time of the kinetics experiments. Data for
JM-216 and -221 were not collected above pH 7.5 since biphasic
kinetic traces, attributed to hydrolysis of the platinum()
compounds, hampered the data analysis. Constant temperature
( 0.1 ЊC) was maintained by an external RM6 LAUDA
circulating-water bath for stopped-flow measurements and a
CARY Peltier thermostat ( 0.01 ЊC) coupled with a circulating-
water temperature control unit for measurements using the
Cary UV/VIS spectrophotometer. The temperature dependence
of the second-order overall rate constants k was investigated
at pH 6.28 for JM394 and JM576 and 7.12 for JM216 and
JM221. At these pH, Asc2Ϫ dominates the reaction completely
(vide infra).
Chart 1
JM216 and JM221 with glutathione were hampered by early
interference from substitution reactions at the platinum()
product. In keeping with this observation, del Socorro Murdoch
et al. recently described substitution reactions between gluta-
thione/glutathione disulfide and [Pt(en)2Cl2].18 Qualitatively,
however, the glutathione reductions of the Pt() carboxylate
complexes studied in the present work proceed at similar rates
to those of ascorbate.
Single-exponential kinetic traces were collected in all cases
and the pseudo-first-order rate constants kobs were derived from
an on-line non-linear least squares fit of the absorbance–time
data using Applied Photophysics20 and Cary WinUV Bio21
software packages. A spectrophotometric investigation of the
reduction of JM216 and JM221 by glutathione (GSH) was not
feasible because of interference from slow subsequent processes
presumed to be substitution of chloride by GSH/GSϪ in the
Pt() reduction product.18,22,23
Experimental
Materials and solutions
The compounds JM216, JM221, JM394, and JM576 were
kindly loaned by the Johnson Matthey Technology Centre
(Reading, Berkshire, UK). -Ascorbic acid (Merck) and sodium
hydrogenascorbate (NaHAsc) (ICN Biochemicals Inc.) were
used as received. All other chemicals used were of analytical
grade. Deionized (Millipore) water was boiled and flushed with
argon for ca. 30 min to remove dissolved dioxygen. Acetate,
phosphate, and TRIS buffers containing 3 × 10Ϫ3 mol dmϪ3
Na2H2(edta) were prepared using oxygen-free water. The Na2H2
(edta) was added to the buffers to sequester trace concentrations
of transition metal ions that might catalyse the autoxidation of
ascorbic acid.19 Deoxygenated solutions of hydrogenascorbate
(3–50) × 10Ϫ3 mol dmϪ3, JM394, and JM576 (0.3–1) × 10Ϫ3 mol
dmϪ3 were prepared immediately before use by dissolving
weighed solid samples directly in buffer. Weighed samples of
JM216 and JM221 were dissolved in water to give (0.4–
1.0) × 10Ϫ3 mol dmϪ3 solutions.
Results and discussion
Stoichiometry and reaction products
The stoichiometry [Pt()]:[Asc]tot for ascorbate reduction of
platinum() halide complexes has been observed to be 1:1.24
By analogy, equimolar quantities of the reactants are assumed
to be consumed in the present reductions. Ascorbate is also
known to be oxidised to dehydroascorbate (DHA) by some
transition metal ions and complexes.25 Proton NMR spectra of
the starting materials and the product mixture at 25 ЊC showed
that cis-[PtCl2(cha)(NH3)] (JM118), cis-[Pt(OAc)2(cha)(NH3)],
and trans-[Pt(OAc)2(cha)(NH3)] were the sole reduction prod-
ucts of JM216, JM394, and JM576, respectively, along with
DHA. The two acetate ligands were released to solution when
1168
J. Chem. Soc., Dalton Trans., 2000, 1167–1172

View Article Online
Table 1 Second-order overall rate constants k as a function of pH for
reduction of JM216, JM221, JM394, and JM576a
k/molϪ1 dm3 sϪ1
Compound
JM216
pH
Measured
Calculated
7.00
7.12
7.25
7.38
7.52
7.00
7.12
7.25
7.38
7.52
4.00
4.25
4.50
4.77
5.00
5.26
5.50
5.98
6.28
6.5
0.083 0.005
0.100 0.003
0.142 0.005
0.18 0.01
0.25 0.01
0.047 0.003
0.059 0.001
0.083 0.001
0.109 0.003
0.17 0.01
0.420 0.004
0.657 0.003
1.13 0.01
1.69 0.02
3.00 0.01
4.79 0.02
8.27 0.04
28.1 0.4
0.078 0.001
0.102 0.002
0.138 0.002
0.186 0.003
0.257 0.004
0.0460 0.0003
0.0600 0.0004
0.0810 0.0005
0.109 0.001
0.151 0.001
JM221
JM394
Fig. 1 Dependence of pseudo-first-order rate constants kobs for reduc-
tion of JM394 (᭺, x = 1), JM576 (᭝, x = 1.5), and JM216 (ꢀ, x = 10)
on [H2Asc]tot. Conditions: pH 7.38, 35 ЊC (ꢀ); pH 5.26, [ClϪ] = 10 mM,
25 ЊC (O, ᭝). For clarity the rate constants for JM216 and JM576 have
been multiplied by x = 10 and 1.5, respectively.
JM216 was reduced; a singlet at δ 2.20 assigned to the co-
ordinated acetates in JM216 disappeared and a new peak at
δ 2.00 due to free acetate was observed. Reduction of JM221 is
assumed to give the same product as that of JM216. No acetate
ligands were released to solution when JM394 and JM576 were
reduced. Two peaks at δ 2.231 and 2.218 assigned to the non-
equivalent acetate groups in JM394 disappeared and a singlet at
δ 2.052 appeared upon reduction. A singlet at δ 2.063 observed
upon reduction of JM576 is assigned to the trans acetate lig-
ands in the platinum() product formed. Proton NMR spectra
of JM-216 and -576 and their reduction products are available
as ESI Fig. S1.
48.5 0.6
90
118
186
230
1
2
4
6
6.66
6.86
7.00
4.50
5.26
6.28
JM576
0.532 0.011
1.92 0.02
21.8 0.2
a Reaction conditions: 25 ЊC, [ClϪ] = 10–20 mmol dmϪ3, 3 mmol dmϪ3
Na2H2(edta) for JM394 and JM576; 35 ЊC, 3 mmol dmϪ3 Na2H2(edta)
for JM216 and JM221. Errors are given as one standard deviation.
Kinetics
Plots of the pseudo-first-order rate constants kobs versus [Asc]tot
at constant pH are linear with zero intercept. Such plots are
illustrated in Fig. 1 for JM216, JM394, and JM576. JM221 is
assumed to behave similarly. Thus, the redox reactions follow
the second-order rate law defined by eqn. (1) where k denotes a
Table 2 Second-order rate constants for reduction of JM216, JM221,
JM394, and JM576 by hydrogenascorbate HAscϪ (k2) and ascorbate
Asc2Ϫ (k3) at 25 ЊC and 1.0 mol dmϪ3 ionic strength
Compound
k2/molϪ1 dm3 sϪ1
k3/molϪ1 dm3 sϪ1
Ϫd[Pt()]/dt = kobs[Pt()] = k[Asc]tot[Pt()]
(1)
JM216
JM221
JM394
JM576
Not observed
Not observed
0.548 0.004
ca. 0.3a
672 15
428 10
pH-dependent second-order overall rate constant. Since there
are no protolytic equilibria associated with the Pt() com-
pounds, under the experimental conditions, the pH dependence
of k is attributed to displacement of equilibria involving the
three protolytic species of ascorbic acid. We then arrive at the
reaction shown in Scheme 1.
(4.46 0.01) × 106
ca. 2 × 106a
a Estimated based on the observation that JM576 is about 2 times less
reactive than JM394 (cf. Table 1).
hydrogenascorbate (HAscϪ) and ascorbate (Asc2Ϫ) are con-
sidered to be the redox active species (Scheme 1).
In eqn. (2), Ka1 and Ka2 are the acid dissociation constants of
ascorbic acid with the values: pKa1 = 3.96 (25 ЊC, 1 mol dmϪ3),26
27
3.90 (35 ЊC, 1 mol dmϪ3
)
and pKa2 = 11.24 (25 ЊC, 1 mol
dmϪ3),28 11.06 (35 ЊC, 1 mol dmϪ3, extrapolated).27 The second-
order rate constants for reduction of JM394 by hydrogen
ascorbate (k2) and by ascorbate (k3) derived from the fit of
eqn. (2) to the experimental data are given in Table 2.
In view of the observation that ascorbate Asc2Ϫ is about 7
orders of magnitude more reactive than hydrogenascorbate
HAscϪ it is assumed that the former is the dominant reductant
for JM216 and JM221 in the near neutral pH region studied.
Accordingly, eqn. (2) is reduced to eqn. (3) and fitted to the
Scheme 1
The second-order overall rate constants k for reduction of
JM216 and JM221 calculated using eqn. (1) and those of
JM394 and JM576 obtained as slopes of the plots of kobs vs.
[Asc]tot are collected in Table 1 together with their pH variation.
In deriving eqn. (2) which describes the pH-dependence of k,
k3Ka2
(3)
k =
aH ϩ Ka2
experimental data. The rate constants k3 derived from the curve
fitting are 886 14 and 524 4 molϪ1 dm3 sϪ1 for reduction at
35 ЊC of JM216 and JM221, respectively. Values recalculated to
25 ЊC are given in Table 2.
k2Ka1aH ϩ k3Ka1Ka2
(2)
k =
aH2 ϩ Ka1aH ϩ Ka1Ka2
J. Chem. Soc., Dalton Trans., 2000, 1167–1172
1169

View Article Online
Table 3 Second-order overall rate constants and activation parameters for the reduction of JM216, JM221, JM394, and JM576 by ascorbate,
Asc2Ϫ a
Compound
JM216
T/ЊC
k/molϪ1 dm3 sϪ1
∆H‡ b/kJ molϪ1
∆S‡ b/J KϪ1 molϪ1
20
25
30
35
40
25
30
35
40
10
15
20
25
30
15
20
25
30
35
(3.34 0.09) × 10Ϫ2
(5.08 0.02) × 10Ϫ2
(7.23 0.01) × 10Ϫ2
0.100 0.003
0.140 0.007
(3.25 0.05) × 10Ϫ2
(4.4 0.1) × 10Ϫ2
(5.94 0.05) × 10Ϫ2
(8.3 0.2) × 10Ϫ2
14.0 0.2
21.8 0.1
32.7 0.6
48.5 0.6
73.2 0.2
8.4 0.2
14.4 0.3
21.8 0.2
34.3 0.6
52
46
1
1
Ϫ97
4
JM221
JM394
Ϫ120
Ϫ24
4
2
56.2 0.5
JM576
63
2
Ϫ8
5
50.1 0.6
a Conditions: pH 7.12, 3 mmol dmϪ3 Na2H2(edta) for JM216 and JM221; pH 6.28, [ClϪ] = 10 mmol dmϪ3, 3 mmol dmϪ3 Na2H2(edta) for JM394 and
JM576. Errors are given as one standard deviation. b Activation parameters refer to reduction with ascorbate, Asc2Ϫ, i.e. k3 pathway in Scheme 1.
Reaction mechanism
Most studies on ascorbic acid reduction of platinum() com-
plexes have involved chloro and hydroxo compounds where
outer-sphere and platinum() catalysed inner-sphere mechan-
isms have been proposed.29 However, to date there has been no
detailed mechanistic investigation on the reduction of plat-
inum() dicarboxylate compounds with ascorbic acid. Since
platinum() compounds are generally substitution inert,8
a
substitution-controlled inner-sphere mechanism is unlikely.
Based on the results of the kinetic studies and on the nature
of the reduction products, an outer-sphere mechanism is sug-
gested for reduction of compounds JM216 and JM221 and a
chloride-bridged reductive elimination mechanism for JM394
and JM576. The latter involves a reductive attack by the ascor-
bate anion Asc2Ϫ on co-ordinated chloride leading to the form-
ation of a bridged activated complex of the type formulated for
JM576 (cf. Chart 2).24 Despite the electronegativity of chloride,
Chart 2
Fig. 2 Plots of the second-order overall rate constants k as a function
of activity of oxonium ions for JM216 (᭝) and JM221 (ꢀ) (a) and as a
function of pH for JM394 (᭺) (b). The solid lines represent the fits of
eqn. (2) and (3) to the experimental data.
this entity bound to the highly oxidising Pt() centre could
have appreciable Clϩ character and therefore be susceptible
to a reductant. This inner-sphere mechanism, which is well-
known for reductions of platinum() complexes with other
reductants,10–12,15 leads in the present case simply to a concerted
two-electron transfer from ascorbate to the Pt() centre and
delivers the Pt() product, dehydroascorbate and two elim-
inated chloride ions as products. Such intramolecular paths are
inherently faster than outer-sphere bimolecular paths in
keeping with the present observations.30
The second-order overall rate constants k calculated from
eqn. (3) are in a good agreement with those obtained experi-
mentally (cf. Table 1). This observation supports the assump-
tion that ascorbate is the primary reductant despite the fact that
hydrogenascorbate constitutes more than 99% of the total con-
centration of ascorbic acid in the region 7.0 ≤ pH ≤ 7.5. The
fits of eqn. (2) and (3) to the experimental data are depicted in
Fig. 2.
The contribution of the k2 pathway to the overall reduction
of JM394 at pH 6.28 is only 1%. Hence, the activation param-
eters presented in Table 3 are associated with the reductions of
JM394 and JM576 by Asc2Ϫ. Similarly, the activation param-
eters determined for reductions of JM216 and JM221 refer to
the k3 pathway.
For JM216 and JM221, the chloride-bridged pathway is
energetically unfavourable, since the chloride ligands are co-
ordinated trans to ammonia and cyclohexylamine which are
firmly bound to the metal centre. In addition, the carboxylate
ligands in these complexes are not efficient bridging groups in a
reductive trans elimination process since Pt() is less able to
ϩ
generate CO2 character in the carboxylate for the ascorbate
ion to access.
1170
J. Chem. Soc., Dalton Trans., 2000, 1167–1172

View Article Online
JM394 and JM576 are more than 3 orders of magnitude
more reactive than JM216 which is attributed to the fact that
inner-sphere electron transfer reactions are generally faster
than outer-sphere ones.30 Further compelling evidence for the
proposed chloride-bridged reductive elimination mechanism
for reduction of JM394 and JM576 where the chloride ligands
are co-ordinated trans to each other is the identification of
trans-[Pt(OAc)2(cha)(NH3)] as the product of reduction of
JM576. A reductive elimination reaction of JM576 via an
outer-sphere mechanism would be expected to result in the
release of the inherently more labile acetate ligands instead of
chloride. This property is gauged from the fact that dissociation
of acetate from JM576 takes place slowly in the presence of
chloride at room temperature (cf. ESI NMR spectra, Fig. S2).
In this context it might be interesting to compare the present
results with those recently reported by Ranford et al. for reduc-
tion with cysteine and methionine of cis,trans,cis-
[PtCl2(OAc)2(NH3)2] as a model for JM216.17 In these cases
also, reduction results in release of two acetates per platinum,
and the reduction is much slower than in our previous model
systems.10,11 A reductive elimination via oxygen-bridged elec-
tron transfer is suggested in these two cases,17 but in view of
the present results, an outer-sphere reduction might also be
operative.
Our results pertaining to the reductions of JM216 and
JM221 do not support some of those reported by Choi et al.16
who effectively did not control their pH conditions. Our
experimental conditions allow us to determine the kinetic and
thermodynamic parameters with very good accuracy. We have
maintained a constant pH using buffers with no observable
interference with the reacting systems. Since the apparent
reducing agent for JM216 and JM221 is ascorbate Asc2Ϫ, which
is about seven orders of magnitude more reactive than hydro-
gen ascorbate HAscϪ, failure to control pH will introduce quite
large errors in the kinetic measurements. Moreover, the activ-
ation parameters reported by Choi et al.16 are unrealistic with
extraordinary and unusual uncertainties.31 Further, their con-
clusion that JM221 is reduced twice as fast as JM216 does not
agree with common experience that steric blocking usually
retards the rates of bimolecular processes.32 Nor is it compatible
with our finding that JM216 reacts faster than the sterically
more hindered JM221. This result, on the other hand, does
agree with the report that the rate of reduction of 1,2-
diaminocyclohexanedicarboxylato(oxalato)platinum() com-
plexes to give oxaliplatin decreases as the carboxylate chain
length increases.33
Fig. 3 Isokinetic relationship34 for JM216, JM221, JM394, and
JM576 including data for the model platinum() complexes 1–5 (see
text) from ref. 24. Complexes on the line are subject to reductive elimin-
ation by Asc2Ϫ attack on co-ordinated halide, whereas JM216 and
JM221 which deviate from the relationship are suggested to react by an
outer-sphere mechanism.
mechanism.10–12,15 Using the data in Tables 1–3, the half-life for
reduction of JM216 with 5 mmol dmϪ3 total concentration of
ascorbic acid (15-fold excess) at pH 7.40 and 35 ЊC is calculated
to be ca. 12 min, and that of JM221 ca. 20 min. Thus, reduction
of JM216 and JM221 in a biological medium is fairly rapid
compared to hydrolytic bio-transformation pathways. Reduc-
tions of JM216 and JM221 with glutathione at pH 7.40 also
take place at similar rates to those of ascorbate reduction, but
the kinetics are complicated by parallel substitution processes
in the platinum() products.
Acknowledgements
Financial support from the Swedish Natural Science Research
Council (NFR), a grant for a visiting professorship for A. M. S.
from the Royal Swedish Academy of Sciences, and a grant from
the Swedish International Development Cooperation Agency
(SIDA) within the SAREC sandwich programme for K. L. is
gratefully acknowledged. We are thankful to Johnson and
Matthey Technology Centre for the generous loan of the JM
compounds.
References
1 L. R. Kelland, B. A. Murrer, G. Abel, C. M. Giandomenico, P.
Mistry and K. R. Harrap, Cancer Res., 1992, 52, 822.
2 C. F. J. Barnard, F. I. Raynaud and L. R. Kelland, in Topics in
Biological Inorganic Chemistry, ed. M. J. Clarke and P. J. Sadler,
Springer-Verlag, Heidelberg, 1999, vol. 1, pp. 45–71.
3 K. R. Harrap, Cancer Res., 1995, 55, 2761.
4 M. J. McKeage, M. Jones, P. M. Goddard, M. Valenti, B. A. Murrer
and K. R. Harrap, Cancer Res., 1993, 53, 2581.
5 P. A. Andrews and S. B. Howell, Cancer Cells, 1990, 2, 35.
6 P. Mistry, L. R. Kelland, S. Y. Loh, G. Abel, B. A. Murrer and K. R.
Harrap, Cancer Res., 1992, 52, 6188.
7 J. J. Roberts, R. J. Knox, F. Friendlos and D. A. Lydall, in
Biochemical Mechanisms of Platinum Antitumor Drugs, ed. D. C. H.
McBrien and T. F. Slater, IRL Press Ltd., Oxford, 1986, pp. 29–64.
8 W. R. Mason, Coord. Chem. Rev., 1972, 7, 241.
9 E. Rotondo, V. Fimiani, A. Cavallaro and T. Ainis, Tumori, 1983,
69, 31; E. E. Blatter, J. F. Vollano, B. S. Krishnan and J. C.
Dabrowiak, Biochemistry, 1984, 23, 4817; A. Eastman, Biochem.
Pharmacol., 1987, 36, 4177; L. Pendyala, A. V. Arakali, P. Sansone,
J. W. Cowens and P. J. Creaven, Cancer Chemother. Pharmacol.,
1990, 27, 248; J. L. van der Veer, A. R. Peters and J. Reedijk, J. Inorg.
Biochem., 1986, 26, 137.
Further support for our mechanistic assignments is given by
the isokinetic relationship shown in Fig. 3.34 The data for
JM394 and JM576 agree very well with those derived from a
series of model platinum() complexes, which are reduced by
Asc2Ϫ in reductive elimination processes involving attack on co-
ordinated halide, viz. cis-[PtCl4(NH3)2] (1), trans-[PtCl4(NH3)2]
(2), trans-[PtCl2(en)2]2ϩ (3), [PtCl6]2Ϫ (4), and [PtBr6]2Ϫ (5).24 The
present data for JM216 and JM221, on the other hand, deviate
significantly from the isokinetic plot, implying reduction by a
different mechanism, i.e. in this case an outer-sphere process.
The unfavourable entropies of activation for JM216 and JM221
(Table 3) might reflect the requirement for substantial solvation
of the leaving carboxylate groups in the activated complex in
these instances.
10 T. Shi, J. Berglund and L. I. Elding, Inorg. Chem., 1996, 35, 3498.
11 T. Shi, J. Berglund and L. I. Elding, J. Chem. Soc., Dalton Trans.,
1997, 2703.
Conclusion
Ascorbic acid reduction of the oral anticancer compounds
JM216 and JM221 in a near neutral aqueous perchlorate
medium follows an outer-sphere mechanism where ascorbate
Asc2Ϫ is the predominant reductant. These two compounds are
reduced at comparable rates but more than 1000 times slower
than JM394 and JM576, whose reduction is proposed to take
place by the usual halide bridged reductive elimination
12 K. Lemma, T. Shi and L. I. Elding, Inorg. Chem., 2000, 39, in the
press.
13 G. K. Poon, F. I. Raynaud, P. Mistry, D. E. Odell, L. R. Kelland,
K. R. Harrap, C. F. J. Barnard and B. A. Murrer, J. Chromatogr. A,
1995, 712, 61; F. I. Raynaud, P. Mistry, A. Donaghue, G. K. Poon,
L. R. Kelland, C. F. J. Barnard, B. A. Murrer and K. R. Harrap,
Cancer Chemother. Pharmacol., 1996, 38, 155.
J. Chem. Soc., Dalton Trans., 2000, 1167–1172
1171

View Article Online
1991, 30, 4555; M. J. Sisely and R. B. Jordan, Inorg. Chem., 1992, 31,
2137; N. Kagayama, M. Sekiguchi, Y. Inada, H. D. Takagi and
S. Funahashi, Inorg. Chem., 1994, 33, 1881.
14 M. J. McKeage, G. Abel, L. R. Kelland and K. R. Harrap, Br. J.
Cancer, 1994, 69, 1.
15 K. Lemma, J. Berglund, N. Farrell and L. I. Elding, J. Biol. Inorg.
Chem., 2000, 5, in the press.
26 M. Kimura, M. Yamamoto and S. Yamabe, J. Chem. Soc., Dalton
Trans., 1982, 423.
16 S. Choi, C. Filotto, M. Bisanzo, S. Delaney, D. Lagasee, J. L.
Whitworth, A. Jusko, C. Li, N. A. Wood, J. Willingham, A.
Schwenker and K. Spaulding, Inorg. Chem., 1998, 37, 2500.
17 L. Chen, P. F. Lee, J. D. Ranford, J. J. Vittal and S. Y. Wong, J. Chem.
Soc., Dalton Trans., 1999, 1209.
18 P. del Socorro Murdoch, N. A. Kratochwil, J. A. Parkinson, M.
Patriarca and P. J. Sadler, Angew. Chem., Int. Ed., 1999, 19, 2949.
19 R. F. Jameson and N. J. Blackburn, J. Chem. Soc., Dalton Trans.,
1982, 9; A. J. Holmes and D. L. H. Williams, Chem. Commun., 1998,
1711; R. A. Rickman, R. L. Sorensen, K. D. Watkins and G. Davies,
Inorg. Chem., 1977, 16, 1570; P. Martinez, J. Zuluaga, J. Kraft and
R. van Eldik, Inorg. Chim. Acta, 1988, 146, 9.
20 Applied Photophysics SX-18MV, Sequential Stopped-Flow ASVD
Spectrofluorimeter, Software manual, Applied Photophysics Ltd.,
Kingston Road, Leatherhead, UK.
21 Cary WinUV Bio Package, Varian Australia, 679 Springvale Road,
Mulgrave, Victoria, Australia 3170.
27 D. H. Macartney and A. McAuley, Can. J. Chem., 1981, 59, 132.
28 D. H. Macartney and N. Sutin, Inorg. Chim. Acta, 1983, 74, 221.
29 U. S. Mehrotra, M. C. Agrawal and S. P. Mushran, J. Inorg. Nucl.
Chem., 1970, 32, 2325; D. J. Evans and M. Green, Inorg. Chim. Acta,
1987, 183; R. N. Bose and E. L. Weaver, J. Chem. Soc., Dalton
Trans., 1997, 1797; K. Hindmarsh, D. A. House and R. van Eldik,
Inorg. Chim. Acta, 1998, 278, 32.
30 H. Taube, Electron Transfer Reactions of Complex Ions in Solution,
Academic Press, New York 1970, ch. 2.
31 Experimental values for enthalpies of activation are generally more
accurate than those for entropies of activation since the former are
obtained from slopes and the latter from a long extrapolation to
1/T = 0 in the Eyring plots.
32 R. G. Wilkins, Kinetics and Mechanisms of Reactions of Transition
Metal Complexes, VCH Verlag, Weinheim, 2nd edn., 1991, pp. 237–
238.
22 T. G. Appleton, J. W. Connor, J. R. Hall and P. D. Prenzler, Inorg.
Chem., 1989, 28, 2030.
23 S. J. Berners-Price and P. W. Kuchel, J. Inorg. Biochem., 1990, 38,
305.
24 K. Lemma, D. A. House, N. Retta and L. I. Elding, unpublished
work.
33 R. Kuzu, T. Nakanishi, M. Miyazaki, T. Tashiro, M. Noji, A.
Matsuzawa, M. Eriguchi, Y. Takeda, N. Akiyama and Y. Kidani,
Anti-Cancer Drugs, 1996, 7, 248.
34 For a summary of the significance of such plots, see for instance:
J. H. Espenson, Chemical Kinetics and Reaction Mechanisms,
McGraw-Hill Inc., New York, 2nd edn., 1995, pp. 164–165.
25 See, for example: C. Creutz, Inorg. Chem., 1981, 20, 4449; B. Bänsch,
Paper a909484i
P. Martinez, D. Uribe, J. Zuluaga and R. van Eldik, Inorg. Chem.,
1172
J. Chem. Soc., Dalton Trans., 2000, 1167–1172