Synthesis, cytotoxicity and cucurbituril binding of triamine linked dinuclear platinum complexes

Article abstract of DOI:10.1039/b905112k

The platinum complexes trans-[{PtCl(NH₃)₂} ₂(μ-NH₂(CH₂)₃NH ₂(CH₂)₃NH₂)]³⁺ (CT033) and the corresponding N4-dimethyl linked analogue trans-[{PtCl(NH ₃)₂}₂(μ-NH₂(CH₂) ₃N(Me)₂(CH₂)₃NH₂)] ³⁺ (CT233) have been synthesised, and their cytotoxicity, ability to bind cucurbit[7,8]uril (Q[7,8]) and reaction with cysteine studied. Both platinum complexes show good activity in the L1210 cell line and maintain their activity in the corresponding cisplatin L1210/DDP cell line. However, the N4-dimethyl analogue CT233 is approximately 50-times less active than the CT033 complex. This suggests that the insertion of a positive charge into the linking ligand may not, per se, be responsible for the higher cytotoxicity generally observed for dinuclear platinum complexes linked by polyamines. The upfield shifts of the resonances from the methylene protons in the linking triamine ligand observed in the ¹H NMR spectra of either CT033 and CT233 upon addition of either Q[7] or Q[8] indicate that the cucurbituril is positioned over the linking ligand. However, the results show that the protonated secondary amine in CT033 acts as a barrier to encapsulation, with the Q[7,8] being positioned over only one propyl-arm at a time. Alternatively, the entire triamine linking ligand of CT233 is fully encapsulated within the Q[7,8] cavity. Encapsulation by Q[7,8] was found to reduce the rate of reaction of CT033 and CT233 with the thiol containing amino acid cysteine, with a greater rate reduction observed for CT233. These results are consistent with the NMR results of the Q[7,8] binding studies of the two platinum complexes. For CT033 encapsulated in Q[7,8], one of the two platinum centres is completely exposed to the solvent, whereas, for CT233 both platinum centres are simultaneously positioned within the portals of the cucurbit[n]uril, thereby, affording greater protection.

Full text of DOI:10.1039/b905112k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthesis, cytotoxicity and cucurbituril binding of triamine linked dinuclear
platinum complexes
Yunjie Zhao,a Mark S. Bali,a Carleen Cullinane,b Anthony I. Day*a and J. Grant Collins*a
Received 12th March 2009, Accepted 14th April 2009
First published as an Advance Article on the web 18th May 2009
DOI: 10.1039/b905112k
The platinum complexes trans-[{PtCl(NH3)2}2(m-NH2(CH2)3NH2(CH2)3NH2)]3+ (CT033) and the
corresponding N4-dimethyl linked analogue trans-[{PtCl(NH3)2}2(m-NH2(CH2)3N(Me)2(CH2)3NH2)]3+
(CT233) have been synthesised, and their cytotoxicity, ability to bind cucurbit[7,8]uril (Q[7,8]) and
reaction with cysteine studied. Both platinum complexes show good activity in the L1210 cell line and
maintain their activity in the corresponding cisplatin L1210/DDP cell line. However, the N4-dimethyl
analogue CT233 is approximately 50-times less active than the CT033 complex. This suggests that the
insertion of a positive charge into the linking ligand may not, per se, be responsible for the higher
cytotoxicity generally observed for dinuclear platinum complexes linked by polyamines. The upfield
shifts of the resonances from the methylene protons in the linking triamine ligand observed in the 1H
NMR spectra of either CT033 and CT233 upon addition of either Q[7] or Q[8] indicate that the
cucurbituril is positioned over the linking ligand. However, the results show that the protonated
secondary amine in CT033 acts as a barrier to encapsulation, with the Q[7,8] being positioned over only
one propyl-arm at a time. Alternatively, the entire triamine linking ligand of CT233 is fully
encapsulated within the Q[7,8] cavity. Encapsulation by Q[7,8] was found to reduce the rate of reaction
of CT033 and CT233 with the thiol containing amino acid cysteine, with a greater rate reduction
observed for CT233. These results are consistent with the NMR results of the Q[7,8] binding studies of
the two platinum complexes. For CT033 encapsulated in Q[7,8], one of the two platinum centres is
completely exposed to the solvent, whereas, for CT233 both platinum centres are simultaneously
positioned within the portals of the cucurbit[n]uril, thereby, affording greater protection.
Introduction
Multinuclear platinum complexes, developed by Farrell and co-
workers,1–4 where two or three platinum coordination centres are
linked by aliphatic amine chains, have shown excellent potential
as anti-cancer agents.1–6 While the multinuclear complexes are
highly cytotoxic, and retain their activity in cisplatin resistant
cell lines,2 they have several drawbacks that could limit their
clinical application. They are highly toxic, e.g. clinical trials have
concluded that the maximum tolerated dose (MTD) for the
trinuclear platinum complex BBR3464 is as low as 0.9 mg m-2,
compared to >60 mg m-2 for cisplatin.7 Furthermore, most of the
administered drug binds to thiol containing plasma proteins and
is subsequently degraded into non-active metabolites.
We have proposed that encapsulation of multinuclear complexes
in macrocyclic molecules called cucurbiturils could protect the
platinum complexes from reaction with thiol containing blood
proteins.8 Cucurbiturils (see Fig. 1) are composed of glycoluril
monomer units, and range in size from 5 to 10 units (Q[5]–
Q[10]).9–15 Cucurbiturils have the ability to form inclusion com-
plexes with various molecules, such as diamines.9 The hydrophobic
Fig. 1 (A) The chemical structure of cucurbit[n]uril, where n = 7 and 8,
and (B) the structure of the dinuclear platinum complexes CT033, CT233,
BBR3005, CT008 and BBR3517.
aSchool of Physical, Environmental and Mathematical Sciences, Univer-
sity College, University of New South Wales, Australian Defence Force
Academy, Northcott Drive, Campbell, ACT 2600, Australia. E-mail: a.day@
adfa.edu.au, g.collins@adfa.edu.au; Fax: +61 2 6268 8017
bResearch Division, Peter MacCallum Cancer Centre, Locked Bag No
1, A’Beckett Street, Melbourne, VIC 8006, Australia. E-mail: carleen.
cullinane@petermac.org; Fax: +61 3 9656 1411
cavity of cucurbituril allows the favourable binding of non-polar
sections of a guest molecule, while the carbonyl-rimmed portals
5190 | Dalton Trans., 2009, 5190–5198
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
provide hydrogen bonding potential and favourable electrostatic
interactions for cationic groups.9,16 We have demonstrated that en-
capsulation in Q[7] and Q[8] significantly decreases the rate of the
reaction of trans-[{PtCl(NH3)2}2(m-NH2(CH2)8NH2)]2+ (CT008)
with cysteine.8 For example, after 500 min only 20% of the
platinum complex remains unreacted and intact in the absence
of cucurbituril, whereas with the addition of Q[7] or Q[8], 70%
and 85% of the drug remain intact respectively. From NMR
studies, it was determined that the cucurbituril was positioned
over the entire linking ligand, thereby protecting both platinum
centres from reaction with cysteine. This was most notable for
Q[8] where significant folding of the octamethylene chain was
observed.
Instrumental methods
Elemental analyses (C, H, N) were performed by the Australian
National University Microanalytical Service using a Carlo Erba
1106 Automatic analyser.
NMR spectra were recorded using a Varian Unityplus-400 spec-
trometer and analysed using Varian’s VNMR software. All NMR
experiments where conducted at 25 ◦C unless otherwise stated.
1H NMR spectra were referenced to TSP (0 ppm) at 25 ◦C by
1
using the residual H signal of the respective deuterated solvent.
1D spectra were recorded with between 16 and 1024 transients.
Synthesis of polyamine ligands
While the results of the encapsulation of dinuclear platinum
complexes linked by diamines (such as 1,8-octanediamine in trans-
[{PtCl(NH3)2}2(m-NH2(CH2)8NH2)]2+(CT008)) are encouraging,
it has been demonstrated that complexes linked by polyamines
(e.g. spermidine and spermine) are considerably more cytotoxic
than their diamine counterparts.2,17 However, host–guest bind-
ing studies from the groups of Mock, Kim and Steinke have
shown that a protonated secondary amine provides a barrier to
encapsulation.18–20 This is due to the high-energy cost of placing the
cation inside the hydrophobic cavity of the cucurbituril, most likely
because of the energy cost of desolvating the cation. Consequently,
for a dinuclear platinum complex linked by a triamine, e.g.
dipropylenetriamine, the cucurbituril would bind only over one
propyl group (but “shuttle” between the two propyl-arms), and
thereby, leave one platinum centre exposed to the solvent and
open to attack by the thiol containing proteins in the blood
plasma.
In this study we have examined the binding of Q[7] and
Q[8] to trans-[{PtCl(NH3)2}2(m-NH2(CH2)3NH2(CH2)3NH2)]3+
and the corresponding N4-dimethyl linked analogue trans-
[{PtCl(NH3)2}2(m-NH2(CH2)3N(Me)2(CH2)3NH2)]3+ (see Fig. 1).
The former of these platinum complexes is the triamine
derivative of the prototype dinuclear complex studied by Far-
rell and co-workers, trans-[{PtCl(NH3)2}2(m-NH2(CH2)6NH2)]2+
(BBR3005),2,21,22 while the latter platinum complex allows the
examination of the effect that converting a secondary ammonium
ion to a quaternary ion has on the cucurbituril binding properties.
In addition, the synthesis and cytotoxicity testing of these new
dinuclear platinum complexes provides further information on
the structure–activity relationship for this important class of
drug.
N4-BOC dipropylenetriamine. tert-Butyl bis(3-aminopropyl)-
carbamate was prepared according to the procedure of Laduron
et al.23 In a flask fitted with a Dean–Stark trap with a condenser
N-(3-aminopropyl)propane-1,3-diamine (4.11 g, 31.37 mmol) was
dissolved in 100 mL of 4-methyl-2-pentanone (methyl iso-butyl
ketone, MIBK) under a nitrogen atmosphere (using standard
Schlenk techniques). The mixture was azeotropically distilled until
no further water was produced to RT. 6.84 g (31.37 mmol) of di-
tert-butyl dicarbonate was dissolved in minimum MIBK, and was
added drop-wise over 5 min. After stirring at RT for 30 min,
the solution was extracted with 10 mL of water. The solvent was
removed by rotary evaporation, 50 mL of iso-propanol added
followed by 10 mL of water and the solution stirred at 60 ◦C for
80 min. The bulk solvent was removed by rotary evaporation. The
residue was washed three times with 100 mL of distilled water. The
water phase was combined and removed by rotary evaporation
to yield tert-butyl bis(3-aminopropyl)carbamate in high purity
1
1
(>99% by H NMR) (yield: 4.300 g, 58%), H NMR (CDCl3):
d (ppm) 0.75 (t, 4H, CH2), 1.30 (s, 9H, CH3), 1.56 (m, 4H, CH2),
2.58 (t, 4H, CH2).
Synthesis of N4,N4-dimethyldipropylenetriamine
2,2¢-(Iminodipropane-3,1-diyl)bis(1H-isoindole-1,3(2H)-dione).
To 5.3347 g (40.65 mmol) dipropylenetriamine was added 12.042 g
(81.30 mmol) of phthalic anhydride and heated under reflux in a
◦
sand bath at 200 C for 1 h before the condenser was removed.
Heating was continued till the reaction ceased to evolve bubbles
of steam (approx. 2 h). The product was crystallized from ethyl
acetate, then methanol. The product was filtered, washed with
acetone, and air-dried. (Yield: 5.47 g, 37%). Anal. Calc. for
C22H21N3O4: C: 67.51%; H: 5.41%; N: 10.74%. Found: C: 67.10%;
H: 5.36%; N: 10.52%. 1H NMR (CD3OD): d (ppm) 1.78 (q,
4H, CH2), 2.50 (t, 4H, CH2), 3.63 (t, 4H, CH2), 7.65 (m, 8H,
phathalamide).
Experimental
Materials
Potassium tetrachloroplatinate(II), N-(3-aminopropyl)propane-
1,3-diamine, di-tert-butyl dicarbonate were obtained from Aldrich
Chemical Company and used without further purification. Q[7]
and Q[8] were synthesised as previously described,9,10 and based
upon NMR analysis were ≥99% pure. Routine solvents where
purchased from APS Chemicals, and were of UniVar analytical
grade. trans-Diamminedichloroplatinum(II) was obtained from
Aldrich or prepared from potassium tetrachloroplatinate. Dowex
cation and anion exchange resins were obtained from Bio-Rad
Laboratories.
3-Amino-N-(3-aminopropyl)-N,N-dimethylpropan-1-aminium
chloride, dihydrochloride. 2.0 g (5.110 mmol) of 2,2¢-(imino-
dipropane-3,1-diyl)bis(1H-isoindole-1,3(2H)-dione) and 2.8 g
(26.42 mmol) of anhydrous carbonate were added to 160 mL of
toluene. To the mixture was added 1.212 mL (12.764 mmol) of
dimethyl sulfate and heated under reflux overnight. The mixture
was allowed to cool, then 20 mL of H2O was added and the
solution boiled for 30 min. The solvent was evaporated by rotary
evaporation and the residue redissolved in 100 mL of methanol.
The solution was cooled to 4 ◦C then filtered, and 0.50 mL
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5190–5198 | 5191
©

View Article Online
(5.11 mmol) of hydrazine hydrate was◦added. The solution was
refluxed overnight before cooling to 4 C and filtered. 40 mL of
1 M HCl was added and the solution boiled down to ~40 mL. The
remaining solvent was removed to dryness by rotary evaporation,
leaving the crude product. The final product was purified on a
DOWEX cation exchange column (10 mm ¥ 20 cm). The column
was prepared for use by washing with 6 M HCl (250 mL), then
0.5 M HCl (250 mL) and finally distilled water until the pH was
about 6. The crude compound was dissolved in a minimum of
0.2 M HCl and was loaded onto the column. The column was
washed with 500 mL 0.2 M HCl, 500 mL 0.5 M HCl, with the
final product in 0.5 M HCl. The solvent was removed by rotary
evaporation leaving a yellow oily residue. (Yield: 0.900 g, 67%). 1H
NMR (CD3OD): d (ppm) 2.25 (m, 4H, CH2), 3.13 (t, 4H, CH2),
3.19 (s, 6H, CH3), 3,51 (t, 4H, CH2).
to a stirred solution of 0.3783 g (1.25 mmol) of trans-
diamminedichloroplatinum(II) in 10 mL DMF, and stirred for
18 h in the absence of light under a dry nitrogen atmosphere.
The mixture was filtered through a 0.5 mm PETE syringe filter,
and cooled to -20 ◦C in a 1 : 3 NaCl : ice (w/w) bath. 2.4 mL
of a 0.245 M solution (185 mg) of 3-amino-N-(3-aminopropyl)-
N,N-dimethylpropan-1-aminium triflate was added drop-wise
over 10 min. The solution was stirred at -20 ◦C for 3 h, then
at RT for 1 h. The solution was dried in vacuo then washed
with ether : acetone 1 : 1 then dissolved in water and freeze-dried.
The resulting yellow solid was fractionally crystallized from
0.2 M HCl and MeOH, then dried to give the product as a
mixed triflate–chloride salt. (Yield: 67 mg). For analysis, the
-
product was precipitated as the PF6 salt by addition of excess
NH4PF6 in a concentrated aqueous solution. Anal. Calc. for
C8H34N7Cl2F18P3Pt2: C: 8.55%; H: 3.05%; N: 8.72%, Found: C:
9.16%; H: 3.13%; N: 8.40%. The product was then dissolved in
a minimum volume of H2O. 10 g of anion exchange resin was
treated with saturated NaCl solution, and then washed with
excess H2O. The solution containing the platinum compound was
added to the resin and stirred for 10 min. The resin was removed
by filtration and washed with 3 ¥ 10 mL of H2O. The combined
washings were freeze-dried to give CT233 in near quantitative
3-Amino-N-(3-aminopropyl)-N,N-dimethylpropan-1-aminium
triflate. 0.350 g (1.24 mmol) of 3-amino-N-(3-aminopropyl)-
N,N-dimethylpropan-1-aminium chloride, dihydrochloride was
dissolved in a minimum volume of water. Dowex anion exchange
resin was prepared for use by washing it with 150 mL of 3 M
NaOH. The resin was washed with approx. 500 mL water or until
the eluant was ~pH 7.0. The compound was then loaded onto the
resin, and left for 10 min, before washing with 3 ¥ 30 mL volumes
of water. The combined washings were then titrated with 1.0 g of
triflic acid dissolved in 5 mL of water, until pH reached ~9.5. The
compound was freeze-dried. (Yield: 0.345 g, 86%).
1
yield. H NMR (D2O): d (ppm) 2.24 (m, 4H, CH2), 2.80 (t, 4H,
CH2), 3.14 (s, 6H, CH3), 3.44 (t, 4H, CH2).
Q[n] encapsulation of platinum complexes
For Q[7] titrations, a stock solution of Q[7] in D2O (~5 mM) was
added in 20 mL aliquots to a solution containing 0.5 mmol of
the relevant platinum complex in 0.7 mL of D2O. Q[8] titrations
where conducted by dissolving solid portions of Q[8] (~0.3 mg,
0.2 mmol) in 0.7 mL of D2O containing ~0.5 mmol of the relevant
Pt complex. 1D 1H NMR spectra where recorded for each
addition and NOESY and/or COSY spectra where taken at 1 : 1
platinum complex to cucurbituril ratio when assignments were not
straightforward.
Synthesis of platinum complexes
{Bis[trans-diamminechloroplatinum(II)](l-[N-(3-aminopropyl)-
propane-1,3-diamine])} dichloride, hydrochloride (CT033).
CT033 was prepared by a method based on that of Qu and
Farrell.24 0.2015 g (1.186 mmol) of AgNO3 was dissolved in
3 mL of DMF and added drop-wise to a stirred solution of
0.3675 g (1.22 mmol) of trans-diamminedichoroplatinum(II) in
10 mL DMF, and stirred for 18 h in the absence of light under
a dry nitrogen atmosphere. The mixture was filtered through
a 0.5 mm PETE syringe filter, and cooled to -20 ◦C in a 3 : 1
ice : NaCl (w/w) bath. 0.1316 g (0.569 mmol) of tert-butyl bis(3-
aminopropyl)carbamate in 3 mL of DMF was added drop-wise
over 10 min. The solution was stirred at -20 ◦C for 3 h, then at RT
for 1 h, under anhydrous conditions. The solvent was removed in
vacuo, and the product stirred in 30 mL of ether : acetone (1 : 1
v/v) overnight. The supernatant liquid was pippetted off and the
solid dissolved in a minimal quantity of MeOH. 5 mL of conc.
HCl was added and the solution stirred over 72 h. The mixture
was cooled to 4 ◦C overnight and then filtered. The resulting
yellowish solid was washed with 2 ¥ 10 mL volumes of acetone,
then ether, then dried in vacuo to give CT033 as a yellowish solid.
(Yield: 0.342 g 78%). Anal. Calc. for C6H30N7Cl5Pt2: C: 9.39%;
H: 3.94%; N: 12.77%. Found: C: 9.59%; H: 3.83%; N: 12.31%. 1H
NMR (D2O): d(ppm) 2.12 (m, 4H, CH2), 2.80 (m, 4H, CH2), 3.16
(t, 4H, CH2).
Cysteine reaction assays
A stock solution of buffered cysteine in simulated physiological
conditions was prepared from 120 mM NaCl, 2.7 mM KCl,
26.3 mM KH2PO4, 123.6 mM Na2HPO4 and 4 mM L-cysteine in
D2O solution, as previously described.8 The isotope pH effect was
not accounted for. All solutions were gassed with argon. Free or
cucurbituril encapsulated platinum drug sufficient to form 1 mM
solutions was dissolved in 0.7 mL of the stock solution at 37 ◦C and
a 1D 1H NMR spectrum was taken immediately. Further spectra
were taken at regular intervals.
Cytotoxicity assays
The murine leukaemia line L1210 and the cisplatin resistant
subline L1210/DDP were grown in RPMI 1640 medium supple-
mented with 10% fetal bovine serum (CSL Ltd, Australia). The
cells were maintained in a humidified incubator with 5% CO2 in air
at 37 ◦C and were tested routinely for mycoplasma. Cytotoxicity
was determined using growth inhibition assays. The metal complex
was dissolved in warm ultrapure water and then diluted to the
required concentrations in complete medium. Growth inhibition
{Bis[trans-diamminechloroplatinum(II)](l-[3-amino-N-(3-amino-
propyl)-N,N-dimethylpropan-1-ol-aminium])} trichloride (CT233).
CT233 was prepared by a method based on that of the
groups of Rendina25 and Farrell.24 0.3220 g (1.25 mmol) of
AgOTf was dissolved in 3 mL of DMF and added drop-wise
5192 | Dalton Trans., 2009, 5190–5198
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 1 Cytotoxicity (IC50) in the L1210 murine leukaemia cancer cell
line and its cisplatin resistant line L1210/DDP of CT033 and CT233. IC50
is defined as the concentration of complex (mM) required to inhibit cell
growth by 50%. The resistance factor (RF) is defined as IC50 resistant/IC50
sensitive. IC50 values were determined from at least two independent
experiments
results are for a 48 h continuous exposure of the metal complex
in L1210 and L1210/DDP murine leukemia cells as previously
described.26
Results
Complex
L1210
L1210/DDP
RF
Synthesis
Cisplatin
CT033
0.5
0.13
8.6
3.0
0.053
6.9
0.12
3.2
2.4
0.008
13.8
0.9
0.4
0.8
0.2
The synthesis of N4-BOC-dipropylenetriamine is an improvement
of the method utilised by Farrell et al. in the synthesis of N4-
BOC-spermidine (used for BBR3571),27 which requires selective
protection of the primary amines by a protecting group which is
removed and discarded once the secondary amine is protected
by the BOC group. By utilising a new method of selectively
protecting secondary amines in the presence of primary amines,23
the whole reaction takes place in the one reaction vessel. The
relatively cheap solvent methyl iso-butyl ketone replaces the more
expensive ethyl-trifluoroacetate as the primary amine protecting
group, and doubles as the reaction medium. The result is a more
efficient reaction and a higher purity product in higher yield. The
reaction scheme is shown in Fig 2. The BOC protecting group
is removed after the addition of platinum centres to the primary
amines, leaving the free secondary amine. This step ensures that
unwanted platination of the central amine does not occur.
The platination of diamine linkers is well established, usually
following the work of Farrell et al. in his production of the
original dinuclear complexes.24 This method uses silver nitrate
and transplatin to produce a mono-labile platinum–DMF species,
which is then reacted with the appropriate free base. However,
it was found that the silver nitrate–DMF method needed some
modification to be able to accommodate the new quaternary
polyamines. In the neutralised state (required for platination),
the central quaternary ammonium ion carries a charge, which
requires a counter ion. As a nitrate salt, it was found that the
polyamines were not soluble in DMF, which inhibited the use
of the synthetic methods used by Farrell et al. A solution was
found by incorporating aspects of the platination method used by
Woodhouse and Rendina.25 By using the triflate counterion, it was
found that the quaternary polyamines were quite soluble in DMF
solution. The triflate anion was added to the free base by titrating
triflic acid into the hydroxide form of the free base to a pH of ~9.
CT233
BBR3005a
BBR3571a
a The IC50 values for BBR3005 and BBR3571, using equivalent experimen-
tal conditions, were obtained from reference 2.
To alleviate any problems with mixed salts, silver triflate was used
instead of silver nitrate.
Cytotoxicity
The in vitro cytotoxicities of CT033 and CT233 were deter-
mined against the murine leukaemia cell line L1210 and the
corresponding cisplatin resistant cell line L1210/DDP. The re-
sults are summarised in Table 1. As expected, inclusion of an
amine into the parent dinuclear complex trans-[{PtCl(NH3)2}2(m-
NH2(CH2)6NH2)]2+ (BBR3005) to form CT033 significantly en-
hanced the cytotoxicity. Interestingly, CT033 is approximately two
fold less active than that reported for the corresponding dinuclear
complex linked by spermidine (BBR3571).2 CT233, where the
secondary ammonium ion has been converted into a quaternary
ion through the addition of methyl groups, is significantly less
active than CT033, but maintains its activity in the L1210/DDP
cell line.
Cucurbituril binding of CT033
Fig. 3 shows the titration of CT033 with aliquots of Q[7] and Q[8],
with the changes in chemical shift of the resonances from the non-
exchangeable protons summarised in Table 2. The resonances from
all the methylene protons shifted upfield upon addition of either
cucurbituril, with the resonance from the b protons exhibiting
the greatest shift for Q[7] and the a protons for addition of Q[8].
Fig. 2 Synthesis of BOC-dipropylenetriamine. Reaction conditions, one-pot (a) methyl iso-butyl ketone, azeotropic distillation; (b) di-tert-butyl
dicarbonate, RT; (c) H2O, iso-proponol, reflux.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5190–5198 | 5193
©

View Article Online
Fig. 3 1H NMR spectra of the titration of CT033 (0.7 mM in D2O at 25 ◦C) with Q[7] (A) and Q[8] (B). The Q[7,8] to CT033 ratio is indicated.
Table 2 Chemical shifts of the non-exchangeable proton resonances of
CT033 and Q[7]- and Q[8]-encapsulated CT033 at r = 1. A negative value
for the change in chemical shift (Dd) indicates an upfield shift
also be concluded that the Q[7,8] “shuttles” from one propyl-arm
to the other with fast exchange kinetics.
While CT033 binds Q[8] with slow exchange kinetics, a progres-
sive change in chemical shift of the resonances of the free and
bound metal complexes during the titration was noted. It suggests
that, although there is slow exchange between the free and Q[8]
bound platinum complex, there is another form of exchange which
influences the shielding of the guest within the cucurbituril. It is
concluded that this is due to a second binding mode, possibly on
the outside of the Q[8], which forces the platinum complex to shift
position inside the cavity. As this binding mode is in fast exchange,
the average state is observed in the NMR spectrum. This proposal
is supported by the observation that as the concentration of Q[8]
approaches that of CT033, the shielding values approach that of
the r = 1 complex. Further addition of Q[8] past r = 1 does not
give further increases in shielding.
Chemical shift (ppm)
Proton Free CT033
Q[7]–CT033 (Dd)
Q[8]–CT033 (Dd)
a
b
c
3.14
2.11
2.79
2.61 (-0.53)
1.51 (-0.60)
2.39 (-0.40)
2.58 (-0.56)
1.60 (-0.51)
2.58 (-0.21)
At r = 0.5, two sets of resonances are observed for each set of
methylene protons, indicating slow exchange (on the NMR time
scale) between the free and Q[7,8]-encapsulated metal complex.
From the original study by Mock and Shih using diamino-
alkanes,9 and subsequent research by others using a variety of
guest molecules,16,28–30 it has been established that resonances from
guest protons that are located inside the cucurbituril cavity shift
upfield. Furthermore, protons that are positioned near the centre
of the host cucurbituril cavity exhibit the largest upfield shifts. As
the c methylene resonance shows the smallest upfield shift, and
hence, is less deeply positioned in the cucurbituril cavity than the
Cucurbituril binding of CT233
Fig. 4 shows the titration of CT233 with aliquots of Q[7] solution
up to a 1 : 1 molar ratio. All methylene protons in the aliphatic
linking ligand exhibit upfield shifts. Table 3 summarises the shift
changes of all non-exchangeable protons upon full encapsulation.
The largest upfield shifts are seen for the c methylene and d methyl
protons, indicating that Q[7] binds centrally over the quaternary
+
other methylene protons, it can be concluded that the central NH2
group acts as a barrier to encapsulation. As only one major set
of bound resonances are observed throughout the titration, it can
5194 | Dalton Trans., 2009, 5190–5198
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 4 1H NMR spectra of the titration of CT233 (0.7 mM in D2O at 25 ◦C) with Q[7]. The Q[7] to CT233 ratio is indicated, and the structure on the
right reflects the proposed binding mode.
Table 3 Chemical shifts of the non-exchangeable proton resonances of
CT233 and Q[7]- and Q[8]-encapsulated CT233 at r = 1. A negative value
for the change in chemical shift (Dd) indicates an upfield shift
nitrogen, with the a methylenes positioned near the Q[7] portals.
A smaller upfield shift for the b methylenes and almost no shift
for the a methylenes also supports this form of binding. The two
sets of peaks seen at r = 0.5 is indicative of a slow rate of exchange
between free and bound, and high binding affinity.
Fig. 5 shows the titration of CT233 at several ratios of added
Q[8] solution up to a 1 : 1 molar ratio. All methylene protons in the
aliphatic linking ligand exhibit upfield shifts. Table 3 summarises
the shift changes of all non-exchangeable protons upon full en-
capsulation. The largest upfield shifts are seen for the b methylene
and d methyl protons, indicating that Q[8] binds centrally over the
whole molecule. Larger shifts for the a methylenes compared to the
Chemical shift (ppm)
Proton Free CT233
Q[7]–CT233 (Dd)
Q[8]–CT233 (Dd)
a
b
c
2.80
2.23
3.43
3.14
2.76 (-0.04)
1.54 (-0.69)
2.40 (-1.03)
2.22 (-0.92)
2.50 (-0.30)
1.54 (-0.69)
2.79 (-0.64)
2.43 (-0.71)
d
Fig. 5 1H NMR spectra of the titration of CT233 (0.7 mM in D2O at 25 ◦C) with Q[8]. The Q[8] to CT233 ratio is indicated, and the structure on the
right reflects the proposed binding mode.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5190–5198 | 5195
©

View Article Online
Q[7] complex indicates a more complete encapsulation of the entire
molecule. The N-methyl groups exhibit a smaller shift upon Q[8]-
binding, compared to Q[7]-binding, whereas, the a methylenes
show a larger shift upon Q[8]-binding. This indicates that the whole
linking triamine ligand is folding within the Q[8] cavity. This is
consistent with folding previously observed for the octamethylene
linker of trans-[{PtCl(NH3)2}2(m-NH2-(CH2)8NH2)]2+ with Q[8].8
Alternatively, due to the significantly smaller cavity size of Q[7],
little folding can take place. The two sets of peaks seen at r = 0.5
is indicative of a slow rate of exchange between free and bound,
however the broad, gradually shifting nature of both the “free”
and “bound” intermediates (at r = 0.5) indicates some form of
intermediate exchange process.
Reaction of cysteine with CT033 and CT233
Fig. 6 and 7 show the reaction time course of 1 : 4 molar ratios of
CT033 and CT233, respectively, with cysteine in phosphate buffer
at 37 ◦C. In both cases, the percentage of the free, intact, platinum
complex was reduced to 50% (t1/2) in approximately 5 min. The rate
of reaction of both CT033 and CT233 with cysteine is considerably
faster than that previously observed for trans-[{PtCl(NH3)2}2(m-
NH2(CH2)8NH2)]2+ (CT008).8 The increased rate is consistent with
previous studies where insertion of a positive charge into the
linking ligand of a dinuclear complex significantly enhanced the
rate of reaction with DNA.2 However, it was proposed that the
observed rate acceleration for the reaction with DNA was due
to the stronger reversible DNA binding association of the more
highly charged cationic complex.2
Fig. 7 1H NMR spectra of the reaction of CT233 with four equivalents
of cysteine. The reaction was carried out at 37 ◦C in 150 mM PBS. The
methylene resonances at 2.75 ppm (denoted with a *) in the spectrum
of CT233 were used to determine the percentage of free, intact, metal
complex.
Encapsulation in either Q[7] or Q[8] before the addition of
cysteine slowed the nucleophilic reaction with the thiol group,
as shown in Fig. 8 and 9. The results are summarised in Table 4.
Fig. 6 1H NMR spectra of the reaction of CT033 with four equivalents
of cysteine. The reaction was carried out at 37 ◦C in 150 mM PBS. Peaks
denoted with a * in the spectrum of CT033 were used to determine the
percentage of free, intact, metal complex.
Fig. 8 1H NMR spectra of the reaction of CT033 encapsulated in Q[8]
with four equivalents of cysteine. The reaction was carried out at 37 ◦C in
150 mM PBS. The methylene resonances at 1.93 ppm (denoted with a *)
in the spectrum of CT033 were used to determine the percentage of free,
intact, metal complex.
Based on previous similar studies with glutathione,31 the
platinum complex is degraded into the detached amine ligand
and various Pt–cysteine species in a step-wise process. Initially,
the thiol group of one cysteine reacts with one platinum centre,
then after another cysteine reacts with the other platinum centre
the bridging amine ligand detaches. However, due to the overlap
Table 4 Half-life (t1/2), in minutes, of the reaction of CT033 and CT233
with cysteine at 37 ◦C. T1/2 is defined as the time taken for the free intact
platinum complex to reduce in concentration to 50%
t1/2 CT033/min
t1/2 CT233/min
1
Free
5
15
10
5
40
45
of the H NMR resonances, no attempt was made to follow the
Encapsulated in Q[7]
Encapsulated in Q[8]
individual species in the degradation; only the decrease of the free
intact platinum complex was monitored.
5196 | Dalton Trans., 2009, 5190–5198
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
its larger cavity allowing it to move further over the platinum
centre and further away from the central amine. As the largest
shift observed in the Q[8]/CT033 complex is for the a methylene
(closest to the platinum), it is probable that Q[8] provides better
protection, than Q[7], to only one platinum centre of CT033 at a
time, with the other metal centre exposed to the solution. However,
this may result in the other platinum centre being more susceptible
to nucleophilic attack which is consistent with the cysteine reaction
assays. Symmetrical changes in chemical shift for the methylenes of
both propyl-arms were seen at a 1 : 1 Q[n] to complex ratio, which
indicates that the shuttling from one propyl-arm to the other is fast
on the NMR time scale. For a non-symmetrical triamine linking
ligand, the cucurbituril would be predominantly positioned over
only one alkyl-arm. For example, for BBR3571 that contains a
butyl- and propyl-arm on either side of the secondary amine,
cucurbit[n]uril would predominantly bind over the butyl-arm, as
the cucurbituril binding to 1,4-diaminobutane is much stronger
than the corresponding propyl-analogue.32 Furthermore, it has
been established that two cucurbiturils will not “share” an amine
and form a 2 : 1 complex with a triamine.
The results of this study demonstrate that the methylation of
the secondary amine in the triamine linking ligand of a dinuclear
platinum complex achieved the desired outcome of enabling
cucurbituril to bind centrally over the entire molecule. The large
shift seen for the N-methyl groups confirms this conclusion. The
next largest shift was observed for the b methylenes, which may
indicate the same form of folding as previously observed with
trans-[{PtCl(NH3)2}2(m-NH2(CH2)8NH2)]2+ and Q[8],8 where the
b methylenes are closer to the geometric centre of the cavity
than the c methylenes. The binding of Q[8] to CT233 indicates
the platinum centres are further within the cavity than within
Q[7], as shown by the larger shielding of the a methylenes. This
greater protection of the platinum centres presumably leads to the
observed greater reduction in the rate of the reaction of CT233
with cysteine, compared to that with CT033.
Fig. 9 1H NMR spectra of the reaction of CT233 encapsulated in Q[7]
with four equivalents of cysteine. The reaction was carried out at 37 ◦C in
150 mM PBS. The methylene resonances at 1.65 ppm (denoted with a *)
in the spectrum of CT233 were used to determine the percentage of free,
intact, metal complex.
Interestingly, Q[7] and Q[8] encapsulation of CT033 only slowed
the reaction with cysteine by 2–3 fold, whereas, encapsulation of
CT233 slowed the reaction more significantly, with an increase in
t1/2 by 8–9 fold.
Discussion
Both CT033 and CT233 exhibit good cytotoxicity in the L1210
and L1210/DDP cell lines. As expected from the previous studies
of Farrell and co-workers,2 the inclusion of a secondary amine into
the hexamethylene chain of BBR3005 to form CT033 significantly
increases the cytotoxicity. However, the difference in cytotoxicity
between CT033 and BBR3571 is interesting. CT033, which
contains only one less methylene group in the linking chain, is
two-times less active than BBR3571. CT233 is approximately 50-
times less active than CT033, but it is also less active than the
diamine linked BBR3005. It has been proposed that the reason
for the increased cytotoxicity of multinuclear platinum complexes
linked by aliphatic chains containing secondary amines is due to
the combination of the cationic charge and hydrogen bonding
ability of the amine.2,27 The results of this study suggest that the
increased cytotoxicity is not primarily due to the introduction of
the additional positive charge.
The encapsulation of CT233 also represents a relatively rare
example of a cation being positioned deep within the cucurbituril
cavity. The delocalisation of the positive charge (due to the electron
donating methyl groups) and the reduced thermodynamic cost of
removing a quaternary amine rather than a secondary amine from
water, are probable driving forces for the encapsulation.
References
1 N. Farrell, in Advances in DNA Sequence-Specific Agents, ed. G. B. Jones
and M. Palumbo, JAI Press Inc., New Haven, CT, 1998, vol. 3, p. 179.
2 N. Farrell, in Platinum-based Drugs in Cancer Therapy, ed. L. R. Kel-
land and N. P. Farrell, Humana Press, Totowa, NJ, 2000, pp. 321–338.
3 N. Farrell, in Advances in DNA Sequence-Specific Agents, ed. L. H.
Hurley and J. B. Chaires, JAI Press Inc., New Haven, CT, 1996,
vol. 2, p. 187.
4 N. Farrell, Comments Inorg. Chem., 1995, 16, 373.
The results of the cysteine reaction assays showed that Q[n]
encapsulation of CT233 significantly slowed the reaction of the
platinum complex with the thiol containing amino acid, whereas,
encapsulation of CT033 only slowed the reaction by 2–3 fold.
The NMR results of Q[n] binding to CT033 showed that for
5 B. A. J. Jansen, J. von der Zwan, J. Reedijk, H. den Dulk and J. Brouwer,
Eur. J. Inorg. Chem., 1999, 1429.
6 N. J. Wheate and J. G. Collins, Coord. Chem. Rev., 2003, 241, 133.
7 D. I. Jodrell, T. R. J. Evans, W. Steward, D. Cameron, J. Prendiville, C.
Aschele, C. Noberasco, M. Lind, J. Carmichael, N. Dobbs, G. Camboni,
B. Gatti and F. D. Braud, Eur. J. Cancer, 2004, 40, 1872.
8 M. S. Bali, D. P. Buck, A. J. Coe, A. I. Day and J. G. Collins, Dalton
Trans., 2006, 5337.
9 W. L. Mock and N.-Y. Shih, J. Org. Chem., 1986, 51, 4440.
10 A. I. Day, A. P. Arnold, R. J. Blanch and B. Snushell, J. Org. Chem.,
2001, 66, 8094.
+
both Q[7] and Q[8] the central NH2 moiety acts as a barrier
to encapsulation, as indicated by the relatively small shifts in
the neighbouring c methylenes. Interestingly, Q[8] exerts an even
smaller shielding effect on the c methylenes, presumably due to
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5190–5198 | 5197
©

View Article Online
11 C. Marquez, U. Pischel and W. M. Nau, Org. Lett., 2003, 5, 3911.
12 W. Ong, M. Gomez-Kaifer and A. E. Kaifer, Org. Lett., 2002, 4, 1791.
13 H. J. Kim, W. S. Jeon, Y. H. Ko and K. Kim, Proc. Natl. Acad. Sci.
U. S. A., 2002, 99, 5007.
14 J. Lagona, P. Mukhopadhyay, S. Chakrabarti and L. Isaacs, Angew.
Chem., Int. Ed., 2005, 44, 4844.
15 J. Kim, I.-S. Jung, S.-Y. Kim, E. Lee, J.-K. Kang, S. Sakamoto, K.
Yamaguchi and K. Kim, J. Am. Chem. Soc., 2000, 122, 540.
16 Y.-M. Jeon, J. Kim, D. Wang and K. Kim, J. Am. Chem. Soc., 1996,
118, 9790.
22 J. D. Roberts, J. Peroutka and N. Farrell, J. Inorg. Biochem., 1999, 77,
51.
23 F. Laduron, V. Tamborowski, L. Moens, A. Horvath, D. De Smaele
and S. Leurs, Org. Process Res. Dev., 2005, 9, 102.
24 Y. Qu and N. Farrell, Inorg. Chem., 1992, 31, 930.
25 S. L. Woodhouse and L. M. Rendina, Dalton Trans., 2004, 3669.
26 T. Talarico, D. R. Phillips, G. B. Deacon, S. Rainone and L. K. Webster,
Invest. New Drugs, 1995, 17, 1.
27 H. Rauter, R. DiDomenico, E. Menta, A. Oliva, Y. Qu and N. Farrell,
Inorg. Chem., 1997, 36, 3919.
17 A. Hegmans, Y. Qu, L. R. Kelland, J. D. Roberts and N. Farrell, Inorg.
Chem., 2001, 40, 6108.
28 H.-Y. Fu, S.-F. Xue, Q.-J. Zhu, Z. Tao, J.-X. Zhang and A. I. Day,
J. Inclusion Phenom. Macrocyclic Chem., 2005, 52, 101.
29 Y. J. Jeon, S.-Y. Kim, Y. H. Ko, S. Sakamoto, K. Yamaguchi and K.
Kim, Org. Biomol. Chem., 2005, 3, 2122.
18 W. L. Mock and J. Pierpont, J. Chem. Soc., Chem. Commun., 1990,
1509.
19 S. I. Jun, J. W. Lee, S. Sakamoto, K. Yamaguchi and K. Kim,
Tetrahedron Lett., 2000, 41, 471.
20 D. Tuncel and J. H. G. Steinke, Macromolecules, 2004, 37, 288.
21 J. D. Roberts, J. Peroutka, G. Beggiolin, C. Manzotti, L. Piazzoni and
N. Farrell, J. Inorg. Biochem., 1999, 77, 47.
30 N. J. Wheate, D. P. Buck, A. I. Day and J. G. Collins, Dalton Trans.,
2006, 451.
31 M. E. Oehlsen, Y. Qu and N. Farrell, Inorg. Chem., 2003, 42, 5498.
32 M. V. Rekharsky, Y. H. Ko, N. Selvapalam, K. Kim and Y. Inoue,
Supramol. Chem., 2007, 19, 39.
5198 | Dalton Trans., 2009, 5190–5198
This journal is
The Royal Society of Chemistry 2009
©