Synthesis, characterisation and in vitro cytotoxicity studies of a series of chiral platinum(II) complexes based on the 2-aminomethylpyrrolidine ligand: X-ray crystal structure of [PtCl2(R-dimepyrr)] (R-dimepyrr = N-dimethyl-2(R)-aminomethylpyrrolidine)

Article abstract of DOI:10.1016/j.ejmech.2008.12.022

A series of platinum(II) complexes were synthesised based on the enantiomerically pure amino acid proline. Novel synthetic pathways were developed, adapted from standard peptide chemistry, to produce the 2-aminomethylpyrrolidine (pyrr) ligand and its derivatives with differing arrangements of methyl substituents at the exocyclic amine sites. The crystal structure of [PtCl₂(R-dimepyrr)] (R-dimepyrr = N,N-dimethyl-2(R)-aminomethylpyrrolidine) is reported and the five-membered ligand ring has been shown to be in an envelope conformation. Cytotoxicity studies were carried out on the ovarian cancer A2780 tumour cell line and its cisplatin-resistant variant, A2780cisR. Remarkably good activity was seen for several of the drugs when compared to cisplatin despite the addition of substantial steric bulk to the amine groups, and there was a lack of cross-resistance with cisplatin seen for some compounds.

Full text of DOI:10.1016/j.ejmech.2008.12.022

European Journal of Medicinal Chemistry 44 (2009) 2807–2814
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis, characterisation and in vitro cytotoxicity studies of a series of chiral
platinum(II) complexes based on the 2-aminomethylpyrrolidine ligand: X-ray
crystal structure of [PtCl2(R-dimepyrr)] (R-dimepyrr ¼ N-dimethyl-2(R)-
aminomethylpyrrolidine)
,
a,
**
Connie I. Diakos a, Mei Zhang b, Philip J. Beale c, Ronald R. Fenton a , Trevor W. Hambley
*
a School of Chemistry, The University of Sydney, Sydney, NSW 2006, Australia
b Department of Radiation Oncology, Royal Prince Alfred Hospital, Camperdown, NSW 2050, Australia
c Department of Medical Oncology, Royal Prince Alfred Hospital, Camperdown, NSW 2050, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of platinum(II) complexes were synthesised based on the enantiomerically pure amino acid
proline. Novel synthetic pathways were developed, adapted from standard peptide chemistry, to produce
the 2-aminomethylpyrrolidine (pyrr) ligand and its derivatives with differing arrangements of methyl
substituents at the exocyclic amine sites. The crystal structure of [PtCl2(R-dimepyrr)] (R-dimepyrr ¼ N,N-
dimethyl-2(R)-aminomethylpyrrolidine) is reported and the five-membered ligand ring has been shown
to be in an envelope conformation. Cytotoxicity studies were carried out on the ovarian cancer A2780
tumour cell line and its cisplatin-resistant variant, A2780cisR. Remarkably good activity was seen for
several of the drugs when compared to cisplatin despite the addition of substantial steric bulk to the
amine groups, and there was a lack of cross-resistance with cisplatin seen for some compounds.
Ó 2008 Elsevier Masson SAS. All rights reserved.
Received 27 July 2008
Received in revised form
26 November 2008
Accepted 16 December 2008
Available online 25 December 2008
Keywords:
Platinum
Anti-tumor
Cytotoxic activity
Enantioselectivity
Structure activity
1. Introduction
complexes containing the cyclohexane-1,2-diamine (chxn) ligand is
evident from the high activity in cisplatin-resistant L1210 leukemia
Platinum(II) anti-cancer agents are now arguably the most
widely used chemotherapeutic agents in the world, and with their
increasing use in combination therapies through emerging syner-
gies with molecularly targeted agents, their place in the clinic is
only set to strengthen [1–4]. However, one of the major problems
preventing a more wide-spread use of platinum agents in anti-
cancer therapy has been their severe side-effects, leading to
a continuing effort on the part of researchers world-wide to
discover new platinum analogues. Thousands of platinum(II)
compounds have been synthesised and screened in the last 30
years since the discovery of cisplatin [5,6] and more than 28 have
been clinically trialled, but few have reached the clinic [2,6].
Chiral diamine ligands enjoy a long history in the research
towards cytotoxically active platinum(II) drugs [5,7]. The discovery
by Burchenal et al. of the lack of cross-resistance of platinum(II)
lines [8–10] and heralded the start of investigation into chiral
diamines. The work done by Kidani et al. into the trans-1,2-chxn
complexes [11–20] has culminated in the use of the drug oxaliplatin
for the treatment of colorectal cancers [20]. Oxaliplatin is now one
of the top ten selling drugs with world-wide sales in the 2005–
2006 financial year of US$1.6 billion.
The synthesis and anti-tumour activity of the dichlor-
oplatinum(II) complex of the 2(S)-aminomethylpyrrolidine (pyrr)
ligand was first reported by Brunner et al. in 1986 [21] and more
recently, Morikawa et al. described the synthesis of a series of
ligands based on the 2-aminomethylpyrrolidine ligand [22–24].
Platinum(II) complexes of the 2-aminomethylpyrrolidine ligand
were synthesised as both the dichloro and cyclobutane-1,1-dicar-
boxylate (CBDC) derivatives, with the R-enantiomer of the latter
going on to Phase III clinical trials as the drug DWA-2114R and the
S-enantiomer being rejected early on for reasons of toxicity [24–
26]. Since the differences between the activity and toxicity of chiral
complexes arises from interaction with chiral targets, they can
provide valuable insights into the factors that control such
interactions.
* Corresponding author. Tel.: þ61 2 9351 2781; fax: þ61 2 9351 3329.
** Corresponding author. Tel.: þ61 2 9351 2830; fax: þ61 2 9351 3329.
E-mail addresses: r.fenton@chem.usyd.edu.au (R.R. Fenton), t.hambley@
chem.usyd.edu.au (T.W. Hambley).
0223-5234/$ – see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.12.022

2808
C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
We report here the synthesis of a series of platinum(II)
2.2. Preparation of N-carbobenzoxy-proline-p-nitrophenol ester
complexes of enantiomerically pure ligands based on the amino
acid proline (Scheme 1). These chiral pyrrolidine ligands are
methylated derivatives of the parent pyrr ligand that has only
hydrogen atoms at three different substituent sites, two at an
exocyclic nitrogen and one at an endocyclic nitrogen atom. These
ligands have a five-membered ring containing a chiral (stereogenic)
carbon atom, from which extends a methylene bridge to the
exocyclic nitrogen atom. Both the exocyclic and endocyclic nitrogen
atoms can become chiral upon coordination to the platinum(II)
atom. The crystal structure of the platinum(II) complex containing
one of these ligands is also reported here.
The synthesis of the intermediate common to 2-aminomethyl
pyrrolidine, N-methyl-2-aminomethylpyrrolidine, and N-dimethyl-
2-aminomethylpyrrolidine was carried out as follows. Following
the method of Berger et al. [27], benzyl chloroformate (17.39 g,
0.102 mol) and sodium hydroxide (2 M, 65 cm3, Fluka) were added
simultaneously over 15 min with vigorous stirring to an ice-cooled
solution of proline (10.0 g, 0.087 mol) in sodium hydroxide (2 M,
44 cm3). The solution was then extracted with diethyl ether
(2 ꢂ100 cm3) and the aqueous layer retained and acidified to
approximately pH 4 with 6 M HCl. The resulting solution was
extracted with ethyl acetate (2 ꢂ 50 cm3) and the organic layer
washed with water, dried over anhydrous sodium sulfate and
filtered, yielding N-carbobenzoxy-proline. This solution was used in
the following step.
2. Experimental section
2.1. Instrumentation
Following the method of Bodanszky and du Vigneaud [28],
p-nitrophenol in 20% excess (14.52 g, 0.104 mol, Fluka) was added
to a 0.2–0.5 M solution of N-carbobenzoxy-proline in ethyl acetate
and the solution cooled to 0 ꢀC. Dicyclohexylcarbodiimide (DCC)
(17.95 g, 0.087 mol) was added and the solution stirred for half an
hour; it was then allowed to equilibrate to room temperature and
stirred for a further hour.
The solid dicyclohexylcarbodiimide urea that formed during the
reaction was removed by suction filtration and washed with ethyl
acetate. The combined washings and filtrate were taken to dryness
by rotary evaporation. The resulting oil was redissolved in chloro-
form to precipitate further DCC urea that was again removed by
suction filtration. The solution was cooled and the product
precipitated by the slow addition of ethanol. The solid product was
removed under suction and the process repeated to produce
further crops. The crude product was further purified by recrys-
tallisation from hot ethanol to obtain pure N-carbobenzoxy-
proline-p-nitrophenol ester as fine white crystals. Yield: 17.68 g,
55% (S-enantiomer), 22.82 g, 71% (R-enantiomer). 1H NMR: solvent
DMSO-d6, ppm; 1.98 (q, 2H); 2.36 (m, 2H); 3.52 (m, 2H); 4.62 (m,
1H); 5.14 (q, 2H); 7.28 (d, 5H); 7.36 (d, 5H); 8.30 (q, 2H). m.p.:
(93 ꢃ 1) ꢀC.
Melting points were measured on a Gallenkamp digital melting
point apparatus and are reported uncorrected. 1H NMR spectros-
copy of the ligands was carried out on a Bruker Avance 300 MHz
spectrometer. 1H and 13C NMR spectroscopy of the platinum(II)
complexes was carried out on a Bruker AMX 400 MHz spectrom-
eter. All spectra were recorded using commercially available
solvents (Merck) of 99.6% isotopic purity or better. The optical
rotations of the ligands were measured at 589 nm using an Optical
Activity POLAAR 2001 Dual Wavelength Polarimeter with a 1 dm
cell at ambient temperature. Mass spectral analysis of the ligands
was carried out on a Hewlett Packard 5989A Mass Spectrometer
Engine and are presented as a chemical ionisation spectrum. The
sample was introduced through the particle beam interface to the
chemical ionisation source. The reagent gas used was methane,
with a source temperature of 220 ꢀC. The spectra were recorded by
Dr. X.-M. Song. The circular dichroism measurements for the plat-
inum(II) complexes were performed on a JASCO J-710 spec-
tropolarimeter equipped with J-700 software for Windows. The
spectra were recorded using approximately 10ꢁ3 M N,N-dime-
thylformamide solutions between 260 and 500 nm, with a sensi-
tivity setting of 200 mdeg and a spectral band width of 1.0 nm. The
instrument was calibrated prior to use with camphor sulfonate
2.3. Preparation of 2-aminomethylpyrrolidine (pyrr)
(lmax ¼ 291.5 nm,
Q
¼ 188.7 mdeg). Samples were placed in
a cylindrical quartz cell (width 1 cm). Diffuse reflectance infrared
Fourier transform spectra (DRIFTS) of the platinum(II) complexes
were obtained on a BIO-RAD FTS-40 spectrophotometer with Win-
IR Windows software. Potassium bromide was used for the
background and matrix over the range of 400–4000 cmꢁ1. Solvents
used were of laboratory grade and were used without further
purification; THF was dried and stored over sodium wire. Due to the
suspected carcinogenic nature of benzene, care was taken during its
use, including use in a fumehood at all times. Proline was
purchased from Aldrich in its pure S- and R-enantiomeric forms and
was used as obtained. Microanalysis was carried out at the Research
School of Chemistry, Australian National University, Canberra, for
analysis of carbon, hydrogen and nitrogen. The absorbance of
treated A2780 cancer cells was read in a SpectraCount, Packard
microplate reader.
N-Carbobenzoxy-proline-p-nitrophenol
ester
(27.62 g,
0.075 mol) was dissolved in methanol (w800 cm3) and cooled to
0 ꢀC. The solution was saturated with dry ammonia gas, triethyl-
amine (5 cm3) was added, and the mixture left to stand at room
temperature for 2 h. The solution was again cooled to 0 ꢀC, satu-
rated with dry ammonia gas and left to stand at room temperature
for 2 h. This process was repeated a further four times and the
solution was then refluxed for 2.5 h. The solvent was removed by
rotary evaporation to give the product as a white solid contami-
nated with some yellow oil. The oil was removed by redissolving
the residue in chloroform and mixing with neutral activated
alumina (50 g) and filtering under suction through a pad of neutral
activated alumina. The solvent was again removed by rotary
evaporation. This process was repeated, removing as much of the
residual p-nitrophenol as possible; the remainder was then
removed using a short column of neutral activated alumina and
eluting with chloroform. Yield: 16.76 g, 90% (S-enantiomer),
15.46 g, 83% (R-enantiomer).
N-Carbobenzoxy-proline-amide (16.76 g, 0.068 mol) was dis-
solved in methanol (200 cm3), palladium on carbon catalyst was
added and nitrogen was bubbled through the mixture for a few
minutes to displace any oxygen. Hydrogen gas was gently bubbled
through the solution until evolution of carbon dioxide ceased, as
determined by passing the exit gas through a saturated calcium
Scheme 1.

C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
2809
hydroxide solution and monitoring the formation of insoluble
calcium carbonate. The reaction was complete within 7 h. The Pd/C
catalyst was removed by filtration under suction through a pad of
celite and the solvent removed by rotary evaporation to leave the
product as a white solid. The yield for this step was essentially
quantitative.
Quantitative yields were achieved for this step. The compound was
treated with hydrogen gas over Pd/C catalyst as before, having
dissolved the dimethylamide product (15.86 g, 0.057 mol) in
methanol (200 cm3). An essentially quantitative yield was again
obtained.
The dimethylamide product (7.26 g, 0.051 mol) was reduced as
described earlier with LiAlH4 (9.69 g, 0.256 mol) and the mixture
refluxed for 48 h. The crude N-dimethyl-2-aminomethylpyrrolidine
was obtained as a light yellow oil. This was purified by Kugelrohr
distillation and the fraction at 90–125 ꢀC (w0.4 mm Hg) collected to
give the pure product as a clear oil. Yield: 2.61 g, 38% (S-enantiomer),
2.13 g, 31% (R-enantiomer); m/z 128, calc. 128. 1H NMR: solvent
CDCl3, ppm; 1.38 (m,1H); 1.70 (m, 2H); 2.21 (s, 6H); 2.31 (t,1H); 2.88
Following the method of Fenton et al. [29], proline-amide (8.52 g,
0.075 mol) was dissolved in dry tetrahydrofuran (THF, 250 cm3) and
the mixture cooled to 0 ꢀC. Freshly crushed lithium aluminium
hydride (14.16 g, 0.375 mol) was slowly added and the mixture was
left to stir at 0 ꢀC for 1 h then allowed to slowly come to room
temperature. After refluxing for 48 h the reaction was quenched by
adding a solution of water (24 g) in THF (200 cm3), dropwise at first,
until all reaction ceased. The resultant solid was filtered off under
suction and the filter cake washed with hot THF (3 ꢂ 250 cm3). The
combined filtrate and washings were reduced to leave the crude
product as a pale brown oil. This oil was further purified by vacuum
distillation; the fraction at 75–95 ꢀC (w0.4 mm Hg) was collected,
giving the pure 2-aminomethylpyrrolidine as a clear oil. Yield:
1.51 g, 20% (S-enantiomer), 2.27 g, 30% (R-enantiomer); m/z 100,
calc.100.1H NMR: solvent CDCl3, ppm; 0.97 (d,1H); 1.10 (d, 2H); 1.31
(m, 1H); 1.54 (s, H2O); 1.76 (m, 1H); 2.65 (m, 1H); 2.91 (m, 2H); 3.05
(m, 1H); 3.21 (m, 1H). [
a
]D þ 0.9ꢀ {ꢁ4.7ꢀ} (c ¼ 1, methanol).
2.6. Preparation of the platinum(II) complexes
Using a procedure developed by Fenton [33] from a published
method
[34],
cis-dichlorobis(dimethylsulfoxide)platinum(II)
(0.422 g,1 mmol) [35,36] was suspended in methanol (40 cm3). The
appropriate diamine ligand (1 mmol) in methanol (20 cm3) was
added and the mixture stirred at room temperature until all solids
had dissolved, giving a clear, pale yellow solution. The solution was
stirred for a further 1.5 h and the methanol removed by rotary
evaporation at 40 ꢀC. The residue was dissolved in water (20 cm3)
and excess lithium chloride added (w0.5 g) and the solution gently
heated on a steam bath until the volume had reduced to around
10 cm3.
(m, 1H). [
a
]
D ꢁ5.8ꢀ {ꢁ1.0ꢀ} (c ¼ 1, methanol).
2.4. Preparation of N-methyl-2-aminomethylpyrrolidine (mepyrr)
Following the method of Fenton [30], N-carbobenzoxy-proline-
p-nitrophenol ester (20.0 g, 0.054 mol) was dissolved in warm
absolute ethanol (400 cm3). Methylamine (alcohol 33%, 40 cm3)
and then triethylamine (5 cm3) were added to the solution. The
mixture was stirred at room temperature for 12 h and the solvent
removed by rotary evaporation to leave a yellow oil contaminated
with a bright orange solid. The oil was redissolved in chloroform
and filtered to remove as much as possible of the solid contami-
nant. The filtrate was then taken to dryness and the displaced
p-nitrophenol removed by running the solution through three
columns of neutral activated alumina, eluting with chloroform,
giving an essentially quantitative yield.
The products formed as fine yellow crystals that were collected
at the pump and washed with a small amount of ice-cold water,
followed by ethanol and then diethyl ether for drying. The crystals
were air-dried for an hour, the aqueous washings and the mother
liquor were retained for a second crop that was collected in the
same manner. The yields of the complexes synthesised are given in
Table 1 along with the results of the microanalyses.
[PtCl2(R-pyrr)] was synthesised by both the above method and
by the method of Morikawa et al. [24]. In the latter method, the
diamine (0.050 g, 1 mmol) was added to a solution of potassium
tetrachloroplatinate(II) (0.400 g, 1 mmol) in water (w8 cm3). The
solution was stirred at room temperature for 3.5 h. A precipitate
formed that was collected at the pump and washed with a small
amount of ice-cold water and ethanol, and air-dried for 1 h to afford
the complex as a pale yellow powder. Yield: 0.165 g (90%).
The benzyl chloroformate protecting group was removed as
described above. The resulting proline-methyl-amide (7.63 g,
0.062 mol) in dry THF (250 cm3) was reduced using freshly
crushed LiAlH4 (11.75 g, 0.395 mol) to give crude N-methyl-2-
aminomethylpyrrolidine as a pale yellow oil. This was purified by
Kugelrohr vacuum distillation with the fraction at 95–120 ꢀC (w0.4
mm Hg) collected to give the pure ligand as a clear oil. Yield: 2.02 g,
33% (S-enantiomer), 1.78 g, 29% (R-enantiomer); m/z 114, calc. 114.
NMR: solvent CDCl3, ppm; 1.32 (m, 1H); 1.74 (m, 3H); 1.92 (s, NH);
2.7. NMR characterisation
All complexes were characterised by NMR and the results are
given below. The spectra of many of the complexes could not be
unambiguously assigned due to their complexity arising from their
formation as diastereomeric mixtures. This reveals a lack of suffi-
cient steric bulk about the five-membered pyrrolidine ring to
enforce a single specific chirality at the coordinating nitrogen atoms.
[PtCl2(pyrr)]. A single set of signals was observed for these
complexes. This is to be expected as the exocyclic nitrogen in this
compound is not chiral, and so no possibility of diastereomers
exists. 1H NMR: solvent DMF-d7, ppm; 1.80 (m, 2H); 1.91 (m, 1H);
2.06 (m, 1H); 2.59 (m, 1H); 2.68 (m, 1H); 3.07 (m, 1H); 3.33 (m, 2H);
5.46 (d, NH2); 6.49 (s, NH). 13C NMR: solvent DMF-d7, ppm; 25.18
(CH2); 25.25 (CH2); 51.64 (CH2); 52.18 (CH2); 67.43 (CH).
[PtCl2(mepyrr)]. Both enantiomers show evidence of diastereo-
mers in both the complexity of the 1H NMR spectrum and a second
minor set of signals in the 13C NMR spectra. S-enantiomer: 1H NMR:
solvent DMF-d7, ppm; 1.76 (m, 2H); 1.97 (m, 2H); 2.61 (s, 3H); 2.71
(d, 2H); 2.98 (m, 1H); 3.12 (m, 1H); 3.28 (m, 1H); 5.94 (s, NH minor);
6.08 (s, NH major); 6.38 (s, NH minor); 6.47 (s, NH major).
2.40 (s, 3H); 2.51 (t, 2H); 2.88 (m, 2H); 3.18 (m, 1H). [
{ꢁ6.1ꢀ} (c ¼ 1, methanol).
a]
D þ 16.9ꢀ
2.5. Preparation of N,N-dimethyl-2-aminomethylpyrrolidine
(dimepyrr)
N-Carbobenzoxy-proline-p-nitrophenol
ester
(20.0 g,
0.054 mol) was dissolved in warm absolute ethanol (400 cm3).
Dimethylamine hydrochloride (5.30 g, 0.065 mol) was added, fol-
lowed by triethylamine (15 cm3, 0.108 mol). The mixture was
stoppered and left to stir for 12 h, the solution was filtered, and the
solvent removed by rotary evaporation. This afforded the product
as a bright yellow solid that was redissolved in chloroform and
eluted through two columns of neutral activated alumina to
remove the displaced p-nitrophenol. The fractions were observed
to contain large quantities of the solid DCC urea. This was found to
be insoluble in 2-propanol, while the intended product was soluble,
allowing the DCC urea to be removed by filtration under suction.

2810
C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
Table 1
Synthetic results for the platinum(II) complexes.
Complex and molecular formula Yield (%) Calculated (%)
Found (%)
C
H
N
C
H
N
[PtCl2(S-pyrr)], C5H12N2Cl2Pt
[PtCl2(R-pyrr)], C5H12N2Cl2Pt
[PtCl2(S-mepyrr)], C6H14N2Cl2Pt
[PtCl2(R-mepyrr)], C6H14N2Cl2Pt
[PtCl2(S-dimepyrr)], C7H16N2Cl2Pt 70
[PtCl2(R-dimepyrr)], C7H16N2Cl2Pt 70
43
25
100
57
16.40 3.30 7.65 16.70 3.35 6.82
16.40 3.30 7.65 16.32 3.31 7.51
18.96 3.71 7.37 18.89 3.64 7.21
18.96 3.71 7.37 18.89 3.71 7.14
21.33 4.09 7.11 21.50 4.05 7.04
21.33 4.09 7.11 21.24 4.07 6.59
R-enantiomer. 1H NMR: solvent DMF-d7, ppm; 1.76 (m, 2H); 1.99
(m, 2H); 2.61 (s, 3H); 2.73 (s, 2H); 3.01 (m, 1H); 3.14 (m, 1H); 3.30
(m, 1H); 6.08 (s, NH); 6.47 (s, NH).
The 13C spectrum shows that in both enantiomers, the signals
due to the minor diastereomer are slightly less than half as intense
as those from the major diastereomer. S-enantiomer: 13C NMR:
solvent DMF-d7, ppm; 29.3 (CH2, minor); 29.7 (CH2, major); 29.9
(CH2, major); 30.0 (CH2, minor); 45.9 (CH3); 55.5 (CH2, minor); 56.2
(CH2, major); 66.4 (CH2, major); 67.8 (CH2, minor); 68.8 (CH,
major); 69.1 (CH, minor). R-enantiomer: 13C NMR: solvent DMF-d7,
ppm; 29.3 (CH2, minor); 29.7 (CH2, major); 29.9 (CH2, major); 30.0
(CH2, minor); 45.9 (CH3); 55.5 (CH2, minor); 56.2 (CH2, major); 66.3
(CH2, major); 67.8 (CH2, minor); 68.8 (CH, major); 69.1 (CH, minor).
[PtCl2(dimepyrr)]. It was not possible to characterise this
complex by 13C and DEPT NMR spectroscopy as a sufficiently
concentrated solution could not be produced in any of the available
deuterated solvents. 1H NMR: solvent DMF-d7, ppm; 1.28 (s, 1H);
1.65 (m, 1H); 1.82 (m, 1H); 1.91 (m, 1H); 2.01 (m, 1H); 2.62 (dd, 1H);
2.85 (s, 6H); 3.15 (m, 1H); 3.27 (m, 1H); 3.85 (m, 1H); 6.40 (s, NH).
Fig. 1. ORTEP plot for [PtCl2(R-dimepyrr)].
media (Trace Bioscience) supplemented with 10% foetal calf serum
(FCS) and 2 mM glutamine and maintained in an humidified
incubator at 37 ꢀC in 5% CO2. The cells were maintained in expo-
nential growth phase by subculturing routinely with trypsin–EDTA.
Cell counts were performed using a haemocytometer counter
(Weber), with 8 ꢂ 104 and 6 ꢂ 104 A2780 and A2780cisR cells
respectively seeded into each well of flat-bottom 96-well plates
(Corning). Cells were allowed to attach overnight.
2.8. Crystallography
Crystals suitable for X-ray crystallography were obtained by
recrystallisation of the [PtCl2(R-dimepyrr)] complex from hot DMF
followed by slow evaporation. The crystals were mounted on glass
fibres with cyanoacrylate resin. Data was collected using an Enraf
Nonius CAD4 four-circle diffractometer, employing graphite
The complex and cisplatin stock solutions were made up fresh in
RPMI 1640 media at concentrations of 0.5 mM. The cells were
exposed to duplicate serial dilutions of the drug in tissue culture for
monochromated Mo Ka radiation. Cell constants were determined
72 h, at a range of concentrations from 0 to 100 mM. The viability of
by a least-squares fit to the setting angles of 25 independent
reflections. Lorentz, polarisation and absorption corrections were
applied using teXsan software [37]. The structure was solved by
direct methods using SHELXS-86 [38] and refined using least-
squares methods with teXsan. Hydrogen atoms were included at
calculated sites with fixed isotropic thermal parameters. Non-
hydrogen atoms were refined anisotropically. All other calculations
were performed using teXsan. The ORTEP plot [39] (30% thermal
ellipsoids) of the complex is shown in Fig. 1.
the cells following treatment was determined using the MTT (3-[4,
5-dimethylthiazol-2-yl]-2, 5-diphenyl bromide; thiazolyl blue)
cytotoxicity assay [43]. MTT (10
was added to each well and incubated for a further 4 h. The culture
medium was removed from each well, and DMSO (150 L, Sigma)
ml of 5 mg/mL in growth medium)
m
added. The plate was shaken for 5 s and the absorbance measured
immediately at 540 nm in a microplate reader. Activity of the
complexes was measured in terms of IC50 (mM) values, determined
as the drug concentration that reduced the absorbance to 50% of
that in untreated control wells.
2.9. Molecular modelling
Molecular mechanics modelling was performed on a Silicon
Graphics Indigo workstation. The graphics program HyperChem
[40] was used for construction of the models, with strain energy
minimisation carried out using MOMEC-SG [41] with previously
reported force fields [42]. Models of the dichloroplatinum(II)
complexes were minimised until the root mean squared (r.m.s.)
shift in all positional coordinates was less than or equal to 0.001 Å.
3. Results and discussion
3.1. Syntheses
A series of enantiomerically pure ligands derived from the
naturally occurring amino acid proline were produced using
the optically active forms of proline as precursors. It is essential that
the complexes are of high enantiomeric and chemical purity for
them to be meaningful probes of Pt/DNA interactions and this was
confirmed by CD, IR and NMR spectroscopies, and polarimetry. The
pyrr ligand was previously prepared from the naturally occurring
2.10. Cytotoxicity assays
In vitro cytotoxicity assays were performed using cisplatin as
a standard. Activity of the complexes was measured on the human
ovarian cancer cell line A2780 and the cisplatin-resistant variant
A2780cisR. Cells were grown as monolayers in RPMI 1640 culture
a-amino acid S-proline [44,45] and Morikawa and colleagues
detailed a synthetic method based on catalytic hydrogenation of 2-
pyrrolcarboxaldoxime [22]. Substituted diamines were obtained by

C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
2811
catalytic hydrogenation of the appropriate 2-alkylaminome-
thylpyrrole [23]. In both cases, the ligands were obtained as racemic
mixtures and were used without resolution into their separate
enantiomers in the synthesis of the platinum(II) complexes and the
subsequent biological studies. Later methods published by the
same workers were based on synthetic procedures using the chiral
precursor, the amino acid proline [24], with the pure optical
isomers being used in studies from that point onwards.
The synthetic schemes used here are derived from standard
peptide chemistry, involving protection of the pyrrolidine nitrogen,
formation of a suitable leaving group on the acid residue of the
amino acid, substitution reaction and deprotection to form a chiral
amide (Scheme 2). Subsequent reduction of the amide produced
the appropriate diamine.
the amino acid residue using dicyclocarbohexyldiimide as
a coupling agent which has the effect of eliminating water from the
reaction and forcing the equilibrium towards the product side,
forming dicyclocarbohexyl urea as a by-product (Scheme 2(a)).
3.1.2. N-Substituted aminomethylpyrrolidines
Detailed in Scheme 2(b) is the general synthetic scheme fol-
lowed for the synthesis of 2(S)-aminomethylpyrrolidine and the
subsequent ligands in the series. Taking a methanol or ethanol
solution of cbz-p-NO2 ester, triethylamine was used as a non-
substituting, volatile organic base during the addition of the
required amine. For synthesis of the parent amine, gaseous
ammonia was needed, and multiple cycles of bubbling the gas
through the solution until saturation followed by heating were
carried out, with the yield for these ligands, not surprisingly, lower
than in the other substituted amines. Removal of the carbobenzoxy
protecting group was achieved by bubbling hydrogen gas through
methanol solutions of the amides over a palladium on charcoal
catalyst. This method is known to produce a cleaner product than
an alternative method that employs HBr in acetic acid [46]. The
3.1.1. N-Carbobenzoxy-S-proline-p-nitrophenol ester
Synthesis of this compound involved protection of the
secondary nitrogen amine with benzochloroformate, that upon
coupling is known as the carbobenzoxy protecting group, cbz. The
leaving group p-nitrophenol was attached to the carboxyl group of
a
O
C
O
O
C
protection
esterification
C
O
NO2
N
H
N
N
OH
OH
benzyl
chloroformate/
NaOH
p-nitrophenol/
DCC
cbz
cbz
proline
ester intermediate
b
O
C
O
C
O
C
deprotection
H2, Pd/C cat.
amination
O
NO2
N
N
amine/
amine
N
amine
cbz
triethylamine
cbz
H
reduction
LiAlH4/THF
amine
N
H
c
O
protection
esterification
O
C
O
C
C
O
NO2
N
N
ethylchloroformate
benzene/NaOH
1. nitrophenol
N
OH
p-
OH
H
2. DCC/ethyl acetate
C
O
C
O
O
O
CH2
CH3
CH2
CH3
amination
O
C
NH3/triethylamine/
MeOH
N
NH3
reduction
C
O
O
N
NH2
LiAlH4/THF
CH3
CH2
CH3
Scheme 2.

2812
C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
Table 2
resultant amides were then aggressively reduced to the diamines
with lithium aluminium hydride and the diamines purified by
Kugelrohr vacuum distillation.
Crystal data for [PtCl2(R-dimepyrr)].
Crystal
[PtCl2(R-dimepyrr)]
Formula weight M
Formula
Crystal system
Space group
a (Å)
394.22
PtC7H16N2Cl2
Orthorhombic
P212121
8.309(5)
8.665(5)
15.198(8)
1066(2)
2.456
736
4
0.71073
13.564
0.456
0.264
3.1.3. Platinum(II) complex synthesis
The preparation of the platinum(II) complexes first involved the
synthesis of the platinum(II) intermediate, cis-[PtCl2(DMSO)2], with
which the ligand was reacted. The diamine pyrrolidine ligands
were coordinated to the platinum(II) ion by reaction with equiva-
lent amounts of the cis-[PtCl2(DMSO)2] complex in methanol.
Complete conversion from the chloro–DMSO–platinum(II) inter-
mediate to the dichloroplatinum(II) product following addition of
lithium chloride was monitored by infrared (IR) spectroscopy. The
absence of peaks attributed to the DMSO group provided conclusive
evidence of formation of only the dichloro product.
The [PtCl2(R-pyrr)] complex was initially synthesised via the cis-
[PtCl2(DMSO)2] intermediate. A subsequent attempt at synthesising
this complex presented difficulties in the precipitation of the
complex and the method of Morikawa et al. [24] was used instead.
This latter method, in this case, gave a pure product, in approxi-
mately 35% yield.
b (Å)
c (Å)
V (Å3)
Dcalc (g cmꢁ3
)
F(000)
Z
l
(Å)
m
(mmꢁ1
)
Tmax (mmꢁ1
)
)
Tmin (mmꢁ1
Temperature (ꢀC)
qmax
22.0
27.46
1963
1688
Reflections collected
Observations
Range of hkl
Maximum shift
R (F0)
ꢁ1 / 10; ꢁ1 / 11; ꢁ1 / 19
0.0073
0.037
0.032
Rw
3.2. Characterisation of the platinum complexes
Examination of the angles about the platinum atom shows
constraints caused by the bidentate ligand leading to a compression
of the N(2)–Pt–N(1) angle to 84.3ꢀ. The strain has been borne slightly
more by the exocyclic side of the ligand, with the Cl(2)–Pt–N(2) angle
having opened up to 93.3ꢀ. This strain is also reflected in the Pt–N
bond lengths. The exocyclic amine Pt–N(2) bond length is an
unusually long 2.066(8) Å, while the endocyclic amine Pt–N(1) bond
length is a more normal 2.02(1) Å. The crystal structure of (ꢁ)-(R)-[2-
(aminomethyl)pyrrolidine](1,1-cyclobutanedicarboxylato)platinum-
(II) monohydrate (DWA-2114R), containing the parent ligand has
a much shorter Pt–N (exocyclic) bond length of 1.998(5) Å, while the
Pt–N (endocyclic) bond length, at 2.019(4) Å, is similar to that found
here [48]. The five-membered ring of the pyrrolidine ligand adopts
an envelope conformation with the C(2) atom deviating by more
than 0.3 Å from the approximately planar moiety comprising the four
atoms N(1), C(1), C(4) and C(3). The C(4) carbon atom is a stereogenic
centre, and here has R chirality. The endocyclic nitrogen, N(1), is also
chiral, and it too has R chirality.
The complexes were characterised for chemical and optical
purity using 1H, 13C and DEPT NMR and CD spectroscopy respec-
tively. In all cases the solvent used for these analyses was DMF (for
CD spectra) or DMF-d7. The DMF signals in the 1H spectra, due to an
isotropic impurity, appear at 2.79, 2.94 and 7.90 ppm; in the 13C
NMR spectra two multiplets appear at 30 and 35 ppm. The reso-
nances due to these impurities were used as internal calibrants.
DMF was chosen as the solvent due to the high solubility of the
complexes in it, and it has been found that the chloride ligands did
not exchange with DMF [47].
The NMR spectra of the complex in which the exocyclic amine
was chiral produced a 13C NMR spectrum that had two sets of
signals and consequently could not be assigned unambiguously.
The multiple signals indicate the formation of diastereomers which
implies a lack of sufficient steric bulk about the five-membered
pyrrolidine ring to enforce a single chirality at the coordinating
nitrogen atoms. Complexes that lacked chirality at the exocyclic
nitrogen atom were always able to be fully assigned because the 13C
NMR spectra of these complexes showed only a single set of signals.
The CD spectra of the complexes confirmed the enantiomeric
purity of the complexes. The four sets of enantiomers produced
spectra with the same qualitative shape with bands in approxi-
mately the same positions.
3.4. Molecular modelling of the Pt(II) complexes
The pyrrolidine system contains three potential chiral sites: the
chiral carbon, plus the endocyclic and exocyclic nitrogen atoms that
may gain a chirality upon coordination to the platinum(II) atom,
depending on the nature of their amine substituent(s). It was found
for each enantiomer that the most energetically favourable struc-
ture was one where the chirality of the carbon and the endocyclic
nitrogen atoms are the same, as is illustrated by the mepyrr
complex (Table 3) and supported by the crystal structure of the
dimepyrr complex.
3.3. Crystallography
The crystal structure of the platinum(II) complex [PtCl2(R-dime-
pyrr)] was determined and is represented in the ORTEP diagram
shown in Fig. 1. Crystallographic data and associated collection
parameters for the structure appear in Table 2. Bond lengths, bond
angles and torsion angles are presented in Tables S1–S3. Lists of the
atomic positional coordinates and thermal parameters of hydrogen
and non-hydrogen atoms and details of least-squares planes calcu-
lations, are included in Tables S4–S7. Observed and calculated
structure factors and intermolecular distances are available on
request.
The [PtCl2(R-mepyrr)] complex has a square-planar geometry
about the platinum(II) atom, with cis chloro ligands and nitrogen
donor atoms. Both the chloro ligands and nitrogen donor atoms
deviate only slightly (0.004–0.005 Å) from the platinum coordination
plane (defined by the platinum and the four donor atoms).
Table 3
Minimised strain energies for the Pt(II) complexes of [PtCl2mepyrr], where Xyz
describes
X carbon chirality, y exocyclic nitrogen and z endocyclic nitrogen
chiralities.
S-Enantiomer
R-Enantiomer
Total strain energy (kJ molꢁ1
)
Sss
Srs
Srr
Ssr
Rrr
Rsr
Rss
Rrs
51.0
52.8
65.6
67.3

C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
2813
Table 4
increasing the steric bulk further at the exocyclic amine site by the
addition of two methyl groups, as in the [PtCl2(dimepyrr)]
compounds. This change also gives rise to enantioselectivity in the
degree of cross-resistance with cisplatin.
For all compounds, higher drug concentrations were required in
the cisplatin-resistant cell lines, a degree of cross-resistance with-
cisplatin. For the parent [PtCl2(pyrr)] compound and the R-
enantiomers of [PtCl2(dimepyrr)] there was only a slight increase,
implying reduced cross-resistance.
Total strain energies for the pyrrolidine platinum(II) dichloro complexes.
S-Enantiomer
R-Enantiomer
Total strain energy (kJ molꢁ1
)
S-Pyrr
R-Pyrr
48.6
51.0
52.8
55.0
Sss-Mepyrr
Srs-Mepyrr
S-Dimepyrr
Rrr-Mepyrr
Rsr-Mepyrr
R-Dimepyrr
It is important to note that the difference between the total
strain energies of the two lower energy diastereomers, which differ
only in the chirality of the exocyclic nitrogen atom, is minimal
indicating that they can be expected to form in similar amounts.
Indeed, the 13C NMR spectra showed the presence of two species,
and one formed in excess over the other by a ratio of approximately
4:3. The total strain energies of each of the pyrrolidine platinum(II)
complexes with the preferred combination of C(4) and endocyclic
amine configurations are presented in Table 4 and reveal that in the
case with a chiral exocyclic amine group (mepyrr) there is almost
equal probability of each diastereomer forming.
The molecular models of the R-enantiomers of the complexes
are presented in Fig. S2. The complexes illustrated in these
molecular models possess a square-planar geometry about the
platinum(II) atom, with the five-membered pyrrolidine ring
adopting a stable envelope conformation. The angles about the
platinum atom vary by little throughout the series: (85.3 ꢃ 1)ꢀ
about the chelate ligand; (91.5 ꢃ1.7)ꢀ on the exocyclic side of the
complex; (89.1 ꢃ2.3)ꢀ between the chloro atoms; and (89.3 ꢃ 3.4)ꢀ
about the endocyclic side of the complex. The constraint placed on
the complex by the chelate ligand, causing a compression of the
angle between the two coordinating nitrogen atoms, is compen-
sated for by an opening up of the angle on the exocyclic side of the
ligand. The angles about the platinum atom of the crystal structure
of [PtCl2(R-dimepyrr)] are within the range of those measured from
the molecular models. Overlaying the molecular model with the
crystal structure of [PtCl2(R-dimepyrr)] shows good agreement
between the two structures (Fig. S3).
4. Conclusions
This series of compounds was designed to provide control over
the number and orientation of the amine protons in the expecta-
tion that this would determine the number of possible hydrogen
bonds between the complex and DNA and influence the activity of
the compounds. The R configuration at the endocyclic amine allows
for hydrogen bond formation, but the S does not [5]. However, the
chirality at this group does not influence the cytotoxicity except in
the bulkiest compounds, and then only marginally. Therefore,
hydrogen bonding between the complex and the DNA does not
appear to be a major determinant of activity in this series.
In ligands in which the exocyclic nitrogen atom had two
different substituents, chirality was also induced at this atom upon
its coordination to platinum. However, molecular mechanics
studies revealed that the complexes are likely to form with both
exocyclic nitrogen atom chiralities and this was confirmed by 13C
NMR studies. This finding implies that there is insufficient bulk
and/or rigidity in the carbon backbone of the ligand to enforce
a specific chirality onto the exocyclic nitrogen atom. The presence
of diastereomers means that the 2-aminomethylpyrrolidine system
fails to meet the criteria for the design of chiral drugs or chiral
probes of drug/DNA interactions.
Cytotoxicity studies against ovarian cancer A2780 cell line and
its cisplatin-resistant variant A2780cisR showed good activity
for all compounds in comparison to cisplatin and a lack of cross-
resistance to cisplatin in some of them. Thus, the amino-
methylpyrolidine framework can be extensively modified without
loss of activity allowing for tuning of the pharmacological proper-
ties. Therefore, this series of complexes is worthy of further
investigation.
3.5. Cytotoxicity studies
The platinum(II) compound DWA-2114R ([Pt(CBDC)(S-pyrr)])
has previously demonstrated high activity against ovarian cancer in
Phase III trials. The choice of an ovarian cancer cell line in this
experiment is therefore most relevant. In order to probe any
possible cross-resistance to cisplatin, a variant of this cell line
which is resistant to cisplatin was also used. The IC50 values
obtained for cisplatin are the same as those obtained in previous
experiments, within error margins (Table 5).
Both enantiomers of the parent compound, [PtCl2(pyrr)],
showed activity similar to cisplatin in the A2780 cell line (within
error margins). Almost identical activity was seen on the addition
of a single methyl group at the exocyclic amine site as in
[PtCl2(mepyrr)], but a substantial decrease in activity was seen on
Acknowledgements
We thank the Australian Research Council and the Sydney
University Cancer Research Fund for financial support.
Appendix. Supplementary data
Tables of crystal data, positional and thermal parameters, bond
lengths and angles, torsion angles and least square planes.
Supplementary data associated with this article can be found in the
online version, at doi:10.1016/j.ejmech.2008.12.022.
Table 5
References
Inhibition of growth (IC50) of A2780 human cancer cell lines determined by the MTT
growth inhibition assay. Standard deviations are in parentheses.
[1] K. Siafaca Technology, Markets and Service Opportunities, Fut. Oncol. 5 (2/3)
(1999) 1041–1071.
[2] E. Wong, C.M. Giandomenico, Chem. Rev. 99 (9) (1999) 2451–2466.
[3] T.W. Hambley, J. Chem. Soc., Dalton Trans. (2001) 2711–2718.
[4] L.R. Kelland, Nat. Rev. Cancer 7 (2007) 573–584.
Drug
A2780, IC50
(m
M)
A2780cisR, IC50
(
m
M)
Resistance factor ratios
Cisplatin
S-Pyrr
R-Pyrr
S-Mepyrr
R-Mepyrr
S-Dimepyrr
R-Dimepyrr
3.1 (1.1)
3.2 (1.6)
3.1 (0.6)
3.5 (0.3)
4.4 (0.6)
10.0 (1.4)
7.2 (2.8)
7.1 (3.2)
4.1 (0.5)
5.8 (1.3)
7.6 (2.7)
13.4 (3.0)
49.0 (9.0)
8.0 (1.6)
2.3
1.3
1.9
2.2
3.0
4.9
1.1
[5] T.W. Hambley, Coord. Chem. Rev. 166 (1997) 181–223.
[6] K.E. Sandman, S.J. Lippard, Methods for Screening the Potential Antitumour
Activity of Platinum Compounds in Combinatorial Libraries, in: B. Lippert (Ed.),
Cisplatin: Chemistry and Biochemistry of a Leading Anticancer Drug, Verlag
Helvetica Chimica Acta, Zurich, 1999, pp. 523–536.
[7] G. Natile, L.G. Marzilli, Coord. Chem. Rev. 250 (11–12) (2006) 1315–1331.

2814
C.I. Diakos et al. / European Journal of Medicinal Chemistry 44 (2009) 2807–2814
[8] J.H. Burchenal, K. Kalaher, T. O’Toole, J. Chisholm, Cancer Res. 37 (1977)
3455–3457.
[29] R.R. Fenton, F.S. Stephens, R.S. Vagg, P.A. Williams, Inorg. Chim. Acta 197
(1992) 233–242.
[9] J.H. Burchenal, K. Kalaher, K. Dew, L. Lokys, G. Gale, Biochimie 60(1978) 961–965.
[10] J.H. Burchenal, K. Kalaher, K. Dew, L. Lokys, Cancer Treat. Rep. 63 (9–10) (1979)
1493–1498.
[11] Y. Kidani, K. Inagaki, S. Tsukagoshi, Gann 67 (1976) 921–922.
[12] T. Tashiro, Y. Kawada, Y. Sakurai, Y. Kidani, Biomed. Pharmacother. 43 (1989)
251–260.
[30] R.R. Fenton, PhD Thesis, Department of Chemistry, Macquarie University,
Sydney, 1990.
[33] R.R. Fenton, W.J. Easdale, H.M. Er, S.M. O’Mara, M.J. McKeage, P.J. Russell,
T.W. Hambley, J. Med. Chem. 40 (1997) 1090–1098.
[34] F.P. Fanizzi, F.P. Intini, L. Maresca, G. Natile, R. Quaranta, M. Coluccia, L. di Bari,
D. Giordano, M.A. Mariggio, Inorg. Chim. Acta 137 (1987) 45–51.
[35] J.H. Price, A.N. Williamson, R.F. Schramm, B.B. Wayland, Inorg. Chem. 11 (6)
(1972) 1280–1284.
[13] Y. Kidani, K. Inagaki, R. Saito, S. Tsukagoshi, J. Clin. Hematol. Oncol. 7 (1) (1977)
197–209.
[14] Y. Kidani, M. Noji, T. Tashiro, Gann 71 (1980) 637–643.
[15] M. Noji, K. Okamoto, Y. Kidani, T. Tashiro, J. Med. Chem. 24 (1981) 508–515.
[16] R.J. Speer, L.M. Hall, D.P. Stewart, H.J. Ridgeway, J.M. Hill, Y. Kidani, K. Inagaki,
M. Noji, S. Tsukagoshi, J. Clin. Hematol. Oncol. 8 (2) (1978) 44–50.
[17] T. Yamashita, J. Hirose, M. Noji, R. Saito, H. Tomida, Y. Kidani, Biol. Pharm. Bull.
16 (10) (1993) 1014–1018.
[36] R.R. Fenton, F.S. Stephens, R.S. Vagg, P.A. Williams, Inorg. Chim. Acta 201
(1992) 157–164.
[37] Texsan, Molecular Structure Corporation, 1985 and 1992.
[38] G.M. Sheldrick, in: G.M. Sheldrick, C. Kruger, R. Goddard (Eds.), SHELXS-86,
Crystallographic Computing, vol. 3, Oxford University Press, Oxford, 1986, pp.
175–189.
[18] K. Inagaki, C. Ninomiya, Y. Kidani, Chem. Lett. (1986) 233–236.
[19] K. Inagaki, Y. Kidani, Inorg. Chem. 25 (1) (1986) 1–3.
[20] L. Pendyala, Y. Kidani, R. Perez, J. Wilkes, R.J. Bernacki, P.J. Creaven, Cancer Lett.
97 (2) (1995) 177–184.
[39] ORTEP, C.K. Johnson, Oak Ridge National Laboratories, Oak Ridge, 1965.
[40] Hypercube, Enraf Nonius, 1985.
[41] T.W. Hambley, MOMEC-SG, Programme for Strain Energy Minimisation, The
University of Sydney, 1991.
[21] H. Brunner, M. Schmidt, H. Schonenberger, Inorg. Chim. Acta 123 (1986) 201–207.
[22] K. Morikawa, M. Honda, K.-I. Endoh, T. Matsumoto, K.-I. Akamatsu, H. Mitsui,
M. Koizumi, J. Pharm. Sci. 79 (8) (1990) 750–753.
[42] T.W. Hambley, Inorg. Chem. 30 (1991) 937–942.
[43] J. Carmichael, W.G. DeGraff, A.F. Gazdar, J.D. Minna, J.B. Mitchell, Cancer Res.
47 (1987) 936–942.
[23] K. Morikawa, M. Honda, K.-I. Endoh, T. Matsumoto, K.-I. Akamatsu, H. Mitsui,
M. Koizumi, Chem. Pharm. Bull. 38 (4) (1990) 930–935.
[44] H. Brunner, M. Schmidt, G. Unger, H. Schonenberger, Eur. J. Med. Chem. 20 (6)
(1985) 509–512.
[24] K. Morikawa, M. Honda, K.-I. Endoh, T. Matsumoto, K.-I. Akamatsu, H. Mitsui,
M. Koizumi, J. Pharm. Sci. 80 (9) (1991) 837–842.
[45] H. Brunner, R. Kroi
b, M. Schmidt, H. Schonenberger, Eur. J. Med. Chem. 21 (4)
(1986) 333–338.
[25] K.-I. Akamatsu, K. Endo, T. Matsumoto, K. Morikawa, M. Koizumi, K. Koizumi,
H. Mitsui, Jpn. J. Cancer Res. 82 (1991) 724–731.
[46] C.L. Weeks, P. Turner, R.R. Fenton, P.A. Lay, J. Chem. Soc., Dalton Trans. 6 (2002)
931–940.
[26] T. Matsumoto, K. Endoh, K. Akamatsu, K. Kamisango, H. Mitsui, K. Koizumi,
K. Morikawa, M. Koizumi, T. Matsuno, Brit. J. Cancer 64 (1991) 41–46.
[27] A. Berger, J. Kurta, E. Katchalski, J. Am. Chem. Soc. 76 (1954) 5552–5554.
[28] M. Bodanszky, V. du Vigneaud, J. Am. Chem. Soc. 81 (1959) 5688–5691.
[47] E.M. Rezler, R.R. Fenton, W.J. Esdale, M.J. McKeage, P.J. Russell, T.W. Hambley, J.
Med. Chem. 40 (1997) 3508–3515.
[48] M. Yonei, H. Murata, M. Itoh, Y. Watanabe, K. Ochi, M. Honda, Y. Nawata, Acta
Crystallogr. C46 (1990) 137–138.