Synthesis and anticancer activity of silver(I)-N-heterocyclic carbene complexes derived from the natural xanthine products caffeine, theophylline and theobromine

Article abstract of DOI:10.1039/c4dt03679d

A new library of silver(I)-N-heterocyclic carbene complexes prepared from the natural products caffeine, theophylline and theobromine is reported. The complexes have been fully characterised using a combination of NMR spectroscopy, mass spectrometry, elemental analysis and X-ray diffraction analysis. Furthermore, the hydrophobicity of the complexes has been measured. The silver(I)-N-heterocyclic carbenes have been evaluated for their antiproliferative properties against a range of cancer cell lines of different histological types, and compared to cisplatin. The data shows different profiles of response when compared to cisplatin in the same panel of cells, indicating a different mechanism of action. Furthermore, it appears that the steric effect of the ligand and the hydrophobicity of the complex both play a role in the chemosensitivity of these compounds, with greater steric bulk and greater hydrophilicity delivering higher cytotoxicity.

Full text of DOI:10.1039/c4dt03679d

Dalton
Transactions
View Article Online
View Journal
PAPER
Synthesis and anticancer activity of silver(I)–
N-heterocyclic carbene complexes derived from
the natural xanthine products caffeine,
theophylline and theobromine†
Cite this: DOI: 10.1039/c4dt03679d
a
a
b
c
Heba A. Mohamed, Benjamin R. M. Lake, Thomas Laing, Roger M. Phillips* and
a
Charlotte E. Willans*
A new library of silver(I)–N-heterocyclic carbene complexes prepared from the natural products caffeine,
theophylline and theobromine is reported. The complexes have been fully characterised using a combi-
nation of NMR spectroscopy, mass spectrometry, elemental analysis and X-ray diffraction analysis. Fur-
thermore, the hydrophobicity of the complexes has been measured. The silver(I)–N-heterocyclic
carbenes have been evaluated for their antiproliferative properties against a range of cancer cell lines of
different histological types, and compared to cisplatin. The data shows different profiles of response when
compared to cisplatin in the same panel of cells, indicating a different mechanism of action. Furthermore,
it appears that the steric effect of the ligand and the hydrophobicity of the complex both play a role in the
chemosensitivity of these compounds, with greater steric bulk and greater hydrophilicity delivering higher
cytotoxicity.
Received 1st December 2014,
Accepted 15th March 2015
DOI: 10.1039/c4dt03679d
www.rsc.org/dalton
Although platinum-based drugs such as cisplatin have been
pivotal in the fight against cancer, severe side-effects and the
Since their isolation in 1991, N-heterocyclic carbenes (NHCs) development of drug resistance necessitate the need for new
have become ubiquitous in organometallic chemistry, particu- organometallic anticancer drugs. As silver is thought to have
Introduction
1
8
1
2
–5
larly in the field of catalysis.
In more recent years, metal– relatively low toxicity and is already used widely in biomedi-
NHCs have shown promise as antimicrobial (silver–NHCs) and cine, it is a judicious choice of metal to study in cancer chemo-
as antitumour (palladium–, copper–, silver– and gold–NHCs) therapy. However, it is imperative that the material (delivery
6
–11
agents.
The antimicrobial properties of silver have been agent) surrounding the silver also has a low toxicity profile.
exploited for centuries, with silver now being incorporated into Xanthine derivatives (Fig. 1) are natural products and are often
several materials such as wound dressings, creams, deodorants found in beverages and foods such as coffee and chocolate. As
and even clothing. The efficacy of a silver-based antibacterial xanthine derivatives contain an imidazole ring, these are
compound appears to be linked to its bioavailability and pro- potential precursors to NHCs. Due to the low toxicity and high
longed release of silver over a long period of time to prevent versatility of these compounds, they are ideal to study in com-
1
2
reinfection. The release rate is linked to the ancillary ligand bination with silver as potential chemotherapeutic com-
and, as NHCs are strong σ-donors, silver–NHCs can have a pounds. Furthermore, these compounds have themselves
slow silver release rate. In addition to antimicrobial activity, been used medicinally and have potential chemotherapeutic
1
9
several studies have demonstrated that silver–NHCs have a effects. Youngs and co-workers have previously prepared a
1
3–17
potential future in the field of cancer chemotherapy.
silver(I)–NHC complex starting from caffeine, which was found
a
School of Chemistry, University of Leeds, Woodhouse Lane, Leeds, LS2 9JT, UK.
E-mail: c.e.willans@leeds.ac.uk
b
Institute of Cancer Therapeutics, University of Bradford, Bradford, BD7 1DP, UK
c
Department of Pharmacy, University of Huddersfield, Queensgate, Huddersfield,
HD1 3DH, UK. E-mail: R.M.Phillips@hud.ac.uk
†
Electronic supplementary information (ESI) available. CCDC 1036995–1037000.
For ESI and crystallographic data in CIF or other electronic format see DOI:
0.1039/c4dt03679d
1
Fig. 1 Xanthine and naturally occurring derivatives of xanthine.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
2
0
to be active against resistant respiratory pathogens. The tion, in which theophylline was reacted with aryl iodide using
same group also added a hydroxyethyl group to the backbone a copper catalyst (CuI, isobutyrylcyclohexanone, Cs CO ). Imi-
2
3
of the NHC, with the silver(I) complex of this ligand being dazolium salts 2b–2d were then prepared by reaction of the
effective against a variety of cystic fibrosis relevant patho- appropriate imidazole with methyl iodide in DMF at reflux,
2
1
gens. Caffeine-based NHC complexes of gold(I) have recently again in a closed system. A hydroxyethyl functionality was
2
2
been reported and their biological properties investigated.
added to the ligand architecture by reaction of theobromine
Herein, we report the synthesis of a family of xanthine-derived with 2-iodoethanol in DMF at reflux, in the presence of a base
2
1
imidazolium salts, prepared from caffeine, theophylline and (Cs CO ), to give imidazole compound 1e. This was sub-
2
3
theobromine. Silver(I)–NHC complexes were prepared from the sequently reacted with methyl iodide under the same reaction
imidazolium salt precursors and evaluated against eight conditions as previously described, to give imidazolium
cancer cell lines.
salt 2e.
Imidazolium salts 2a–2e were fully characterised, and solid-
state structures were obtained for imidazolium salts 2b–2d
see ESI†). In each case, the backbone pyrimidinone of the
imidazolium salt twists slightly out of the plane as defined by
the N(1)–C(2)–N(4) bond. This twisting is most pronounced in
imidazolium salt 2d, with a C(9)–N(4)–C(7)–O(2) torsion
angle of 9.19° (analogous torsion angles in 2b and 2c are
(
Results and discussion
Imidazolium salts 2a–2e were prepared from either caffeine,
theophylline or theobromine (Scheme 1). Methylation of
caffeine to yield the corresponding imidazolium salt 2a is
achieved through reaction with methyl iodide in DMF at
<1°). The greater degree of twisting in 2d presumably arises
2
0
due to the steric constraints imposed by the phenyl
N-substituent.
reflux. Due to the volatility of methyl iodide, we found it
necessary to use a large excess (20 equivalents) and heat in a
closed system. We found that dimethyl sulfate and methyl tosy-
late could also be used as methylating agents to give similar
product yields of the corresponding imidazolium methyl
sulfate or tosylate salts (>70%). However, reaction of caffeine
with other alkylating agents (benzyl bromide, 1-iodobutane)
under the same conditions did not result in formation of an
imidazolium salt, presumably due to the relatively low basicity
of the nitrogen atom of the caffeine. Therefore, theophylline
was employed as the precursor to initially add an alternative
alkyl group, followed by methylation using methyl iodide. Imi-
dazole compounds 1b and 1c were synthesised through reac-
tion of benzyl bromide or 1-iodobutane with theophylline in
acetonitrile at reflux, in the presence of a base (K CO ). Imida-
Reaction of imidazolium salts 2a–2e with AgOAc at room
temperature resulted in formation of complexes of the type
Ag(NHC)OAc 3a–3e (Scheme 2). Complex 3a was synthesised in
2
0
methanol as previously reported. We found it necessary to
use 1 : 1 mixtures of either methanol–dichloromethane (with
ligand precursors 2b, 2c and 2e) or methanol–acetonitrile
(
with ligand precursor 2d) to aid solubility of the imidazolium
salts for the syntheses of complexes 3b–3e.
Silver(I)–NHC complexes 3a–3e were fully characterised, and
solid-state structures were obtained for complexes 3b–3d
(
Fig. 2). In all three cases the geometry around the silver atom
deviates from linearity, with C(2)–Ag(1)–O(3) bond angles of
68.6–171.6° (Table 1). This deviation is consistent with the
structure of complex 3a, which has previously been reported in
1
2
3
zole compound 1d required an Ullmann-type coupling reac-
2
0
the literature. The silver–carbene bond lengths for complexes
c (2.078(3) Å) and 3d (2.068(4) Å) are similar to those for com-
3
plexes 3a (2.067(3) Å) and 3e (2.072(4) Å) which have previously
been reported. The silver–carbene bond length in complex 3b
(
2.052(7) Å), however, is shorter than those observed in other
similar complexes. This is likely due to the flexibility of the
benzyl substituent, hence lower steric effect, allowing closer
Scheme 1 Synthesis of imidazolium salts 2a–2e starting from caffeine,
theophylline or theobromine.
Scheme 2 Synthesis of silver–NHC complexes 3a–3e.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
Table 1 Selected bond lengths (Å) and angles (°) for silver(I)–NHC com-
plexes 3b–3d
Bond lengths (Å)
and angles (°)
3b
3c
3d
Ag(1)–C(2)
Ag(1)–O(3)
C(2)–Ag(1)–O(3)
N(1)–C(2)–N(4)
Ag(1)–Ag(2)
2.052(7)
2.120(5)
171.4(2)
105.2(5)
—
2.078(3)
2.068(4)
2.111(3)
168.94(11)
105.6(3)
—
2.1042(19)
168.72(9)
105.46(18)
3.1931(3)
a possible argentophilic interaction of approximately 3.2 Å
may cause the lengthening of the silver–carbene bond in the
solid-state structure of this complex.
The in vitro cytotoxicity of silver(I)–NHC complexes 3a–3e
was determined using MTT-based assays involving a 96 hour
drug-exposure period. Compounds were tested for their activity
against A375 (malignant melanoma), HCT116 (colorectal carci-
noma), HT-29 (colorectal adenocarcinoma), LN229 (glioblas-
toma), Panc-1 (pancreatic carcinoma), SiHa (grade II,
squamous cell carcinoma cervix), U-87 MG (glioblastoma) and
U-251 (glioblastoma). The results are summarised in Table 2
and Fig. 3. Complexes 3a–3e exhibit moderate cytotoxicity
against the cell lines tested, with IC50 values being in the
micromolar range.
The profiles of response in this panel of cells is different to
that for cisplatin, indicating a different mechanism of action
for silver(I)–NHC complexes. The most cytotoxic silver complex
in general across these cell-lines is 3d (see ESI†), in which the
NHC ligand bears an N-phenyl substituent. The sterically
encumbering phenyl group is likely to exert a stabilising effect
on the silver–NHC bond, which may lead to a slower silver
release rate and enhanced cytotoxicity profile. It is apparent
that addition of a hydroxyethyl group to the backbone of the
ligand improves activity, with complex 3e being on average
Fig. 2 Molecular structures of silver(I)–NHC complexes 3b–3d. Hydro-
gen atoms have been omitted for clarity and ellipsoids are shown at 50%
probability.
2-fold more cytotoxic than complex 3a. This is presumably due
to the solubility of the complex, and its ability to cross the cell
membrane.
As a means to assess the potential of complexes 3a–3e to
2
3
interaction between the carbenic centre and the silver ion. The enter a cell membrane, the hydrophobicity (log P) of the
n
butyl substituent in complex 3c is not expected to impart any five complexes was measured. Accurate amounts of each com-
significant steric constraints on the carbenic carbon. However, pound were dissolved in octanol-saturated water, with an
Table 2 Response of eight cell lines to silver(I)–NHC complexes 3a–3e and cisplatin. Values presented are IC50 µM ± SD (in parentheses) for three
independent experiments
Cell line
3a
3b
3c
3d
3e
Cisplatin
A357
34.5 ± 3.8
26.7 ± 9.9
41.8 ± 9.8
29.2 ± 12.8
31.7 ± 2.8
21.9 ± 3.8
33.7 ± 1.4
54.8 ± 15.7
21.4 ± 7.5
29.6 ± 3.5
29.9 ± 16.5
18.4 ± 12.0
29.7 ± 8.2
15.3 ± 2.6
22.8 ± 5.1
26.7 ± 8.8
27.4 ± 3.9
22.4 ± 4.8
28.5 ± 2.0
46.5 ± 9.8
16.9 ± 1.1
16.4 ± 0.5
14.0 ± 4.3
51.4 ± 17.8
11.5 ± 5.3
19.5 ± 2.3
21.4 ± 6.5
11.2 ± 1.5
7.6 ± 3.2
13.1 ± 3.7
17.6 ± 4.5
14.2 ± 2.5
12.4 ± 1.3
19.0 ± 5.1
20.7 ± 1.5
7.4 ± 1.8
23.5 ± 5.9
14.0 ± 2.2
22.1 ± 8.1
29.6 ± 5.1
1.2 ± 0.3
2.4 ± 0.3
0.6 ± 0.1
0.7 ± 0.5
2.6 ± 0.9
0.9 ± 0.3
0.9 ± 0.4
1.0 ± 0.6
HCT116
HT-29
LN229
Panc-1
SiHa
U87MG
U-251
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Conclusion
Synthetic procedures have been developed for the preparation
of imidazolium salts from caffeine, theophylline and theo-
bromine. Due to the relatively low basicity of the nitrogen donor
in caffeine, it was found necessary to initially add varying sub-
stituents to theophylline. Following this, methylation was
achieved through heating the imidazole compounds with a
large excess of methyl iodide in a closed system. The imidazo-
lium salts were reacted with silver acetate to prepare silver(I)–
NHC complexes of the type Ag(NHC)OAc. These were fully
characterised and the solid-state structures determined
through the use of X-ray diffraction techniques. The antiproli-
ferative activities of the silver complexes were examined
Fig. 3 Response of eight cell lines to silver(I)–NHC complexes 3a–3e
and cisplatin. Values presented are IC50 (µM) ± SD for three independent
experiments. The presentation of results is based on the sensitivity of against eight cancer cell lines, and compared to cisplatin.
cells to cisplatin with the most sensitive line on the far left whereas the Whilst these complexes were less active than cisplatin against
most resistant line is on the far right. If the test compounds have a
similar mechanism of action to cisplatin, similar patterns of chemosensi-
range. The hydrophobicity of each complex was also measured
tivity would be expected.
the cell lines tested, their IC50 values were in the micromolar
to assess their potential to cross the cell membrane. Upon
comparing the chemosensitivity data for each complex and
considering their hydrophobicity/hydrophilicity profiles, it is
evident that a combination of both steric effects of the ligand
and solubility of the silver complex play a role in their activity
against cancer cells. Sterically bulky N-substituents on the
ligand may result in a slower release rate of silver from the
complex, leading to increased cytotoxicity, whilst a more
hydrophilic complex is likely to have increased penetration
through the cell membrane.
equal amount of water-saturated octanol being layered on top.
This was placed in the vibrax machine for 4 hours at 500 g
−
1
min . The layers were separated and the water-saturated
octanol layer was analysed using UV-vis spectroscopy, with an
average of at least six runs being taken. The results are sum-
marised in Table 3 and S9 in the ESI,† with the complexes
ranging from hydrophilic (−1.96) to hydrophobic (0.56). Cis-
platin, with a log P value of −2.36, is more hydrophilic than
any of the silver(I)–NHC complexes tested. Complex 3e, like cis- Experimental
platin, is hydrophilic, with the other complexes being more
hydrophobic. Hydrophilicity is likely to lead to improved pene-
General considerations
Manipulations were performed under an atmosphere of dry
nitrogen by means of standard Schlenk line techniques. The
gas was dried by passing through a twin-column drying appar-
tration of the cell membrane by complex 3e, hence its
increased cytotoxicity when compared to complex 3a. It is
clear, however, that hydrophobicity/hydrophilicity is not the
single most important factor when considering the effect of
these types of complexes against cancer cells. Though complex
2 5
atus containing molecular sieves (4 Å) and P O . Anhydrous
solvents were prepared by passing the solvent over activated
alumina to remove water, copper catalyst to remove oxygen and
molecular sieves to remove any remaining water, via the Dow-
3
d is more hydrophobic than complex 3a, 3d is on average over
2
-fold more cytotoxic than 3a. This indicates that a combi-
1
13
1
Grubbs solvent system. H and C{ H} NMR spectra were
recorded on a Bruker DPX300 spectrometer. The values of
chemical shifts are given in ppm and values for coupling con-
stants (J) in Hz. Mass spectra were collected on a Bruker Dal-
tonics (micro TOF) instrument operating in the electrospray
mode. Microanalyses were performed using a Carlo Erba
Elemental Analyser MOD 1106 spectrometer. X-ray diffraction
data were collected on an Agilent SuperNova diffractometer
fitted with an Atlas CCD detector with Mo-Kα radiation (λ =
nation of both steric and solubility factors should be taken
into account when designing new NHC ligands for study in
chemotherapeutic applications.
Table 3 Log P values for complexes 3a–3e
0
.71073 Å) or Cu Kα radiation (λ = 1.5418 Å). Crystals were
Complex
Cisplatin
Log P
±SD
mounted under oil on glass or nylon fibres. Data sets were cor-
rected for absorption using a multiscan method, and the struc-
tures were solved by direct methods using SHELXS-97 and
−2.36
−0.01
0.53
0.56
0.1512
−1.96
0.01
0.03
0.02
3
3
3
3
3
a
b
c
d
e
2
0.07 refined by full-matrix least squares on F using ShelXL-97,
0.05
0.06
interfaced through the program X-Seed. Molecular graphics for
all structures were generated using POV-RAY in the X-Seed
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
+
13 4 2
program. Imidazolium salts 2a and 2e and silver(I)–NHC calcd for C13H N O [M + H] : 257.1039. Found: 257.1031.
complexes 3a and 3e were prepared using slightly modified Anal. Calcd for C H N O : C, 60.93; H, 4.72; N, 21.86. Found:
1
3
12 4 2
2
0,21
literature procedures.
Synthesis of 9-benzyl-1,3,5-trimethylxanthine (1b). Theo-
phylline (5.0 g, 27.8 mmol) was dissolved in acetonitrile Imidazole 1b (2.6 g, 9.6 mmol) was dissolved in dimethyl-
80 mL) and benzyl bromide (16.5 mL, 138.7 mmol) and pot- formamide (3 mL) in an ampoule, and methyl iodide
C, 60.60; H, 4.85; N, 21.80.
Synthesis of 9-benzyl-1,3,5-trimethylxanthinium iodide (2b).
(
assium carbonate (4.25 g, 30.8 mmol) were added. The (11.9 mL, 191 mmol) was added. The mixture was heated at
mixture was heated to reflux for 24 hours. The mixture was fil- 70 °C for 24 hours in a closed system resulting in a clear red
tered and washed with acetonitrile. The filtrate was dried solution. Excess diethyl ether was added to the flask to precipi-
in vacuo to yield the product as a white solid. Yield: 6.1 g tate an orange powder which was filtered. Further recrystallisa-
1
(
81%). H NMR (300 MHz, CDCl ) δ: 7.58 (s, 1H, NCHN), tion using dichloromethane–diethyl ether and acetonitrile–
3
7
2
.43–7.32 (m, 5H, aromatic), 5.52 (s, 2H, CH ), 3.60 (s, 3H, diethyl ether resulted in the product as a yellow solid. Yield:
1
3
1
1
3 3 3 6
CH ), 3.13 (s, 3H, CH ). C{ H} NMR (75 MHz, CDCl ) δ: 153.4 0.8 g (20%). H NMR (300 MHz, DMSO-d ): δ 9.49 (s, 1H,
(
CvO), 150.1 (CvO), 139.4 (C), 136.1 (NCHN), 133.6 (C), NCHN), 7.46–7.37 (m, 5H, aromatic), 5.75 (s, 2H, CH ), 4.16 (s,
2
1
1
3
1
27.2, 126.9, 126.6 (CH), 105.1 (C), 48.3 (CH
3
), 31.8 (CH
3
), 27.9 3H, CH
[M + H] : 271.1195. (75 MHz, DMSO-d
Found: 271.1188. Anal. Calcd for C H N O ·1/3Et O: C, 139.4 (NCHN), 134.4 (C), 128.8 (CH), 128.7 (CH), 128.2 (CH),
3
), 3.73 (s, 3H, CH
3 3
), 3.26 (s, 3H, CH ). C{ H} NMR
+
+
(
CH
2
). HRMS (ESI ): calcd for C14
H
15
N
4
O
2
6
): δ 153.1 (CvO), 150.1 (CvO), 139.8 (C),
1
4
14
4
2
2
6
2.43; H, 5.92; N, 18.99. Found: C, 63.00; H, 5.45; N, 19.40.
Synthesis of 9-butyl-1,3,5-trimethylxanthine (1c). Theophyl- (ESI ): calcd for C30
line (5.0 g, 27.8 mmol) was dissolved in acetonitrile (70 mL), 569.2619. Calcd for C H N O Na [M − H + Na] : 307.117096.
106.9 (C), 51.1 (CH
3
), 37.2 (CH
H N O [2M − H] : 569.2614. Found:
33 8 4
3
), 31.3 (CH
3
), 28.4 (CH
2
). HRMS
+
+
+
1
5
16 4 2
4 2 2
and butyl iodide (12.0 mL, 105.5 mmol) and potassium Found: 307.116546. Anal. Calcd for C15H17IN O ·1/4Et O:
carbonate (4.25 g, 30.8 mmol) were added. The mixture was C, 44.61; H, 4.56; N, 13.01. Found: C, 44.80; H, 4.30; N, 12.70.
heated to reflux for 48 hours, filtered, and washed with aceto- MP: 200.2–203.4 °C.
nitrile. The filtrate was dried in vacuo to yield the product as a
Synthesis of 9-butyl-1,3,5-trimethylxanthinium iodide (2c).
-DMSO): Imidazole 1c (4.5 g, 19.1 mmol), methyl iodide (23.7 mL,
δ 8.09 (s, 1H, NCHN), 4.26 (t, J = 7.22 Hz, 2H, CH ), 3.42 (s, 3H, 381 mmol) and dimethylformamide (10 mL) were added to an
1
white solid. Yield: 4.7 g (74%) H NMR (300 MHz, d
6
2
CH
3
), 3.19 (s, 3H, CH
3
), 1.74 (quin, J = 7.22 Hz, 2H, CH
), 0.85 (t, J = 7.22 Hz, 3H, CH
H} NMR (75 MHz, CDCl ): δ 155.1 (CvO), 151.7 (CvO), yellow solid. The solid was filtered and recrystallised from
2
), 1.24 ampoule and heated at 70 °C for 24 hours in a closed system.
1
3
(
sext, J = 7.22 Hz, 2H, CH
2
3
).
C
Excess diethyl ether was added to the solution to precipitate a
1
{
3
1
48.9 (C), 140.8 (NCHN), 106.9 (C), 47.0 (CH
2 2
), 32.8 (CH ), 29.7 dichloromethane–diethyl ether, and dried in vacuo to yield the
+
1
(CH ), 27.9 (CH ), 19.5(CH ), 13.4 (CH ). HRMS (ESI ): calcd product as a bright yellow solid. Yield: 6.3 g (87%). H NMR
2
3
3
3
+
for C H N O [M + H] : 237.1351. Found: 237.1345. Anal. (300 MHz, d -DMSO): δ 9.41 (s, 1H, NCHN), 4.45 (t, J = 7.25 Hz,
1
1
17
4
2
6
Calcd for C11
C, 55.40; H, 6.85; N, 23.30.
H
16
N
4
O
2
: C, 55.92; H, 6.83; N, 23.71. Found: 2H, CH
CH
Synthesis of 9-phenyl-1,3,5-trimethylxanthine (1d). Theo- 2H, CH ), 0.91 (t, J = 7.25 Hz, 3H, CH ). H NMR (300 MHz,
2
), 4.16 (s, 3H, CH
3 2
), 1.82 (quin, J = 7.25 Hz, 2H, CH ), 1.31 (sext, J = 7.25 Hz,
3
), 3.76 (s, 3H, CH
3
), 3.26 (s, 3H,
1
2
3
phylline (4.0 g, 22.2 mmol), copper(I) iodide (1.9 g, CDCl
3
): δ 10.50 (s, 1H, NCHN), 4.58 (t, J = 7.20 Hz, 2H, CH
2
),
1
0.0 mmol), 2-isobutyrylcyclohexanone (2.2 mL, 13.1 mmol) 4.48 (s, 3H, CH
3
), 3.87 (s, 3H, CH ), 3.40 (s, 3H, CH ), 1.99
3
3
and caesium carbonate (6.52 g, 20.0 mmol) were placed in a (quin, J = 7.20 Hz, 2H, CH ), 1.45 (sext, J = 7.20 Hz, 2H, CH ),
2
2
1
3
1
Schlenk flask and degassed. Anhydrous dimethylsulfoxide 0.98 (t, J = 7.20 Hz, 3H, CH
3
). C{ H} NMR (75 MHz, CDCl
3
):
(17 mL) and iodobenzene (3.5 mL, 31.3 mmol) were added to δ 153.1 (CvO), 149.9 (CvO), 139.5 (C), 138.6 (NCHN), 107.8
the flask. The mixture was heated at 130 °C for 24 hours. The (C), 49.7 (CH ), 38.9 (CH ), 32.3 (CH ), 32.1 (CH ), 28.7 (CH ),
3
2
2
3
2
+
dark brown solution obtained was dissolved in dichloro- 19.4 (CH
methane and extracted with water three times. The aqueous [M
3
), 13.3 (CH
3
). HRMS (ESI ): calcd for C12
H
19
N
4
O
2
+
− I] : 251.1508. Found: 251.1589. Anal. Calcd for
layers were combined and washed with dichloromethane three C H IN O : C, 38.11; H, 5.06; N, 14.81. Found: C, 37.80;
1
2
19
4 2
times. All of the organic layers were combined and dried H, 5.00; N, 14.60. M.P: 142.4–144.7 °C.
in vacuo to give a dark brown solid. Recrystallisation from
Synthesis of 9-phenyl-1,3,5-trimethylxanthinium iodide (2d).
dichloromethane–diethyl ether yielded the product as a white Imidazole 1d (0.2 g, 0.78 mmol), methyl iodide (0.91 mL,
solid. To ensure complete removal of the copper, the solid was 14.6 mmol) and dimethylformamide (4 mL) were added in an
dissolved in dichloromethane and a saturated solution of ampoule. The mixture was heated at 70 °C for 24 hours in a
EDTA was added. The organic layer was extracted and dried closed system. Excess diethyl ether was added to the resulting
1
in vacuo to give a white solid. Yield: 1.78 g (34%). H NMR orange solution to precipitate the product as a yellow solid
1
(
7
{
1
1
300 MHz, d
6
-DMSO): δ 8.31 (s, 1H, NCHN), 7.71–7.14 (m, J = that was filtered and dried in vacuo. Yield: 0.15 g (97%). H
1
3
.54 Hz, 5H, aromatic), 3.46 (s, 3H, CH ), 3.18 (s, 3H, CH ).
C
NMR (300 MHz, d -DMSO): δ 9.75 (s, 1H, NCHN), 8.17 (broad
6
3
3
1
H} NMR (75 MHz, d
6
-DMSO): δ 153.7 (CvO), 150.9 (CvO), s, 5H, aromatic), 4.25 (s, 3H, CH
3
), 3.82 (s, 3H, CH
). C{ H} NMR (75 MHz, d -DMSO): δ 154.1 (CvO),
25.1 (CH), 106.1 (C), 29.6 (CH ), 27.8 (CH ). HRMS (ESI ): 151.9 (CvO), 140.3 (C), 139.8 (NCHN), 132.7 (C), 130.7 (CH),
3
), 3.40 (s,
1
3
1
49.2 (C), 142.7 (NCHN), 134.8 (C), 128.9 (CH), 128.5 (CH), 3H, CH
3
6
+
3
3
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
+
1
(
29.3 (CH), 125.8 (CH), 109.5(C), 35.7 (CH
3
), 28.5 (CH
3
), 26.3
28 8 4 2
C H28AgN O [Ag(NHC) ] : 647.1284. Found: 647.1297. Anal.
+
+
CH ). HRMS (ESI ): calcd for C H N O [M − I] : 271.1190. Calcd for C H AgN O : C, 43.96; H, 3.92; N, 12.82. Found: C,
3
14 15
4
2
16 17
4 4
Found: 271.1188. Anal. Calcd for C14
H, 4.68; N, 12.39. Found: C, 36.80; H, 4.20; N, 12.20. M.P:
68.1–169.4 °C.
Synthesis of 9-benzyl-1,3,5-trimethylxanthine-8-ylidene silver In vitro cell tests were performed at the Institute of Cancer
4 2 2
H15IN O ·3H O: C, 37.18; 44.70; H, 4.40, N, 12.60.
Cytotoxicity studies
1
acetate (3b). Imidazolium salt 2b (0.25 g, 0.61 mmol) was Therapeutics, Bradford. Cells were incubated in 96-well plates,
3
added to silver acetate (0.2 g, 1.2 mmol) in a mixture of anhy- at 2 × 10 cells per well in 200 µL of growth media (RPMI 1640
drous dichloromethane (5 mL) and methanol (5 mL). The supplemented with 10% foetal calf serum, sodium pyruvate
mixture was stirred at room temperature for two hours in the (1 mM) and L-glutamine (2 mM)). Cells were incubated for
dark. The mixture was filtered, with the solid being washed 24 hours at 37 °C in an atmosphere of 5% CO prior to drug
2
with dichloromethane, and the filtrate was dried in vacuo to exposure. Silver compounds and cisplatin were dissolved in di-
give the product as a white solid which was recrystallized from methylsulfoxide at a concentration of 25 mM and diluted with
1
dichloromethane–pentane. Yield: 0.3
300 MHz, CDCl
matic), 5.66 (s, 2H, CH
g
(40%).
H NMR medium to obtain drug solutions ranging from 25 µM to
(
3
): δ 7.49 (m, 2H, aromatic), 7.31 (m, 3H, aro- 0.049 µM. The final dimethylsulfoxide concentration was 0.1%
), 4.21 (s, 3H, CH ), 3.79 (s, 3H, CH ), (v/v) which is non-toxic to cells. Drug solutions were applied to
.38 (s, 3H, CH ), 2.05 (s, 3H, CH ). C{ H} NMR (75 MHz, cells and incubated for 96 hours at 37 °C in an atmosphere of
2
3
3
1
3
1
3
3
3
3 2
CDCl ): δ Ag–C not observed, 178.2 (CvO), 153.2 (CvO), 150.7 5% CO . The solutions were removed from the wells and fresh
(
CvO), 140.4 (C), 135.4 (C), 129.1 (CH), 128.8 (CH), 128.5 medium added to each well along with 20 µL MTT (5 mg
−
1
(CH), 109.3 (C), 54.4 (CH ), 40.1 (CH ), 32.1 (CH ), 28.9 (CH ), mL ), and incubated for 4 hours at 37 °C in an atmosphere of
3
3
3
2
+
+
2
6
4
2.4 (CH
3
). HRMS (ESI ): calcd for C30
H
32
N
8
O
4
2 2
Ag [Ag(NHC) ] : 5% CO . The solutions were removed and 150 µL dimethyl-
75.1597. Found: 675.1616. Anal. Calcd for C17
5.25; H, 4.24; N, 12.42. Found: C, 45.30; H, 4.20, N, 12.20.
H
19AgN
4
O
4
: C, sulfoxide was added to each well to dissolve the purple forma-
zan crystals. A plate reader was used to measure the
Synthesis of 9-butyl-1,3,5-trimethylxanthine-8-ylidene silver absorbance at 540 nm. Lanes containing medium only, and
acetate (3c). Imidazolium salt 2c (0.10 g, 0.26 mmol) was dis- cells in medium only (no drug), were used as blanks for the
solved in anhydrous methanol (5 mL) and dichloromethane spectrophotometer and 100% cell survival respectively. Cell
(5 mL) and silver acetate (0.09 g, 0.54 mmol) were added. The survival was determined as the true absorbance of treated cells
mixture was stirred for two hours in the dark. The mixture was divided by the true absorbance of controls and expressed as a
filtered, with the solid being washed with dichloromethane, percentage. The concentration required to kill 50% of cells
and the filtrate was dried in vacuo to yield a white sticky solid. (IC50) was determined from plots of % survival against drug
Recrystallisation from dichloromethane–diethyl ether gave the concentration. Each experiment was repeated 3 times and a
product as a white solid which was further dried in vacuo. mean value obtained.
1
Yield: 0.06 g (55%). H NMR (300 MHz, CDCl
3
): δ 4.44 (t, J =
Hydrophobicity
7
3
1
.35 Hz, 2H, CH
2
), 4.22 (s, 3H, CH
3
), 3.80 (s, 3H, CH ), 3.38 (s,
3
H, CH ), 2.00 (s, 3H, CH ), 1.79 (quin, J = 7.35 Hz, 2H, CH ), Equal volumes of octanol and NaCl-saturated water were
3
3
2
.37 (sext, J = 7.35 Hz, 2H, CH
2
), 0.92 (t, J = 7.35 Hz, 3H, CH
3
). stirred at room temperature for 24 hours, and separated to give
6
-DMSO): δ Ag–C and CvO (acetate) octanol-saturated water and water-saturated octanol. Five stan-
1
3
1
C{ H} NMR (75 MHz, d
not observed, 153.0 (CvO), 150.6 (CvO), 140.8 (C), 108.3 (C), dard concentrations (5, 10, 20, 40 and 60 µM) of the complexes
5
0.3 (CH
3
), 39.4 (CH
2
), 33.0 (CH
3 3 2
), 31.6 (CH ), 28.3 (CH ), were prepared from the octanol-saturated water. Analysis using
+
1
9.10 (CH
2
), 13.6 (CH
3
). HRMS (ESI ): m/z Calcd for UV/vis spectroscopy was used to obtain a calibration curve
+
C H AgN O [Ag(NHC) ] : 608.1910. Found: 608.1939. Anal. of absorbance vs. concentration for each complex at its
2
4
36
8
4
2
Calcd for C14
H
21AgN
4
O
4
: C, 40.30; H, 5.07; N, 13.43. Found: C, maximum absorbance. Accurate amounts of the complexes
3
9.90; H, 5.30, N, 13.10.
Synthesis of
were dissolved in the octanol-saturated water (25 mL) to make
9-phenyl-1,3,5-trimethylxanthine-8-ylidene up a concentration of 50 µM. 3 mL of octanol-saturated water
silver acetate (3d). Imidazolium salt 2d (0.15 g, 0.38 mmol) containing the complex was placed in a centrifuge tube and
was dissolved in anhydrous acetonitrile (5 mL), and methanol 3 mL of water-saturated octanol was layered on top. Six
(
5 mL) and silver acetate (0.125 g, 0.75 mmol) were added. The samples prepared in this manner were shaken for 4 hours
mixture was stirred at room temperature for two hours in the using a vibrax machine at 500 g min− . The layers were separ-
1
dark. The mixture was filtered and the solvent removed from ated and the octanol-saturated water layer was retained for ana-
the filtrate in vacuo to give the product as a white solid. Yield: lysis using UV/vis spectroscopy. The average concentration of
1
0
5
.115 g (69.1%). H NMR (300 MHz, d
6
-DMSO): δ 7.53 (broad s, the six runs was calculated using the calibration graph and
H, aromatic), 4.27 (s, 3H, CH
3
1
), 3.80 (s, 3H, CH ), 3.19 (s, 3H, maximum absorbance for each complex. Subtraction of the
3
1
3
CH ), 1.78 (s, 3H, CH ). C{ H} NMR (75 MHz, d -DMSO): δ average concentration obtained from the concentration of an
3
3
6
1
86.1 (Ag–C), 175.5 (CvO), 152.1 (CvO), 150.5 (CvO), 141.1 unshaken sample in octanol-saturated water gave the final
(
3
C), 137.9 (C), 129.3 (CH), 128.9 (CH), 126.6 (CH), 109.3 (C), [C]org. The [C]org and [C]aq. were used to determine the par-
1.7 (CH ), 28.3 (CH ), 23.2 (CH ). HRMS (ESI ): m/z Calcd for tition coefficient log P.
3 3 3
+
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
1
1 F. Hackenberg and M. Tacke, Dalton Trans., 2014, 43, 8144–
Acknowledgements
8153.
We would like to thank the University of Leeds (CEW, BRML), 12 C. L. Fox and S. M. Modak, Antimicrob. Agents Chemother.,
Leeds International Research Scholarship (HAM) and Institute 1974, 5, 582–588.
of Cancer Therapeutics, Bradford (TL, RMP) for funding. 13 D. A. Medvetz, K. M. Hindi, M. J. Panzner, A. J. Ditto,
We would also like to thank Jonathan Roberts for the TOC
graphic.
Y. H. Yun and W. J. Youngs, Met.–Based Drugs, 2008, 2008,
384010.
14 H. A. Mohamed and C. E. Willans, in Organometallic Chem-
istry: Volume 39, The Royal Society of Chemistry, 2014, pp.
2
6–50.
Notes and references
1
5 S. B. Aher, P. N. Muskawar, K. Thenmozhi and
1
2
3
4
5
6
7
8
A. J. Arduengo, R. L. Harlow and M. Kline, J. Am. Chem.
Soc., 1991, 113, 361–363.
W. A. Herrmann, Angew. Chem., Int. Ed., 2002, 41, 1290–
P. R. Bhagat, Eur. J. Med. Chem., 2014, 81, 408–419.
16 F. Hackenberg, G. Lally, H. Muller-Bunz, F. Paradisi,
D. Quaglia, W. Streciwilk and M. Tacke, J. Organomet.
Chem., 2012, 717, 123–134.
1
309.
V. Dragutan, I. Dragutan, L. Delaude and A. Demonceau, 17 D. C. F. Monteiro, R. M. Phillips, B. D. Crossley, J. Fielden
Coord. Chem. Rev., 2007, 251, 765–794.
and C. E. Willans, Dalton Trans., 2012, 41, 3720–3725.
N. Marion and S. P. Nolan, Acc. Chem. Res., 2008, 41, 1440– 18 B. Rosenber, L. Vancamp and T. Krigas, Nature, 1965, 205,
449.
698–699.
R. Corberan, E. Mas-Marza and E. Peris, Eur. J. Inorg. 19 J. W. Daly, Cell. Mol. Life Sci., 2007, 64, 2153–2169.
1
Chem., 2009, 1700–1716.
W. J. Youngs, A. R. Knapp, P. O. Wagers and C. A. Tessier,
Dalton Trans., 2012, 41, 327–336.
20 A. Kascatan-Nebioglu, A. Melaiye, K. Hindi, S. Durmus,
M. J. Panzner, L. A. Hogue, R. J. Mallett, C. E. Hovis,
M. Coughenour, S. D. Crosby, A. Milsted, D. L. Ely,
C. A. Tessier, C. L. Cannon and W. J. Youngs, J. Med.
Chem., 2006, 49, 6811–6818.
L. Mercs and M. Albrecht, Chem. Soc. Rev., 2010, 39, 1903–
1
912.
M. L. Teyssot, A. S. Jarrousse, M. Manin, A. Chevry, 21 M. J. Panzner, K. M. Hindi, B. D. Wright, J. B. Taylor,
S. Roche, F. Norre, C. Beaudoin, L. Morel, D. Boyer,
D. S. Han, W. J. Youngs and C. L. Cannon, Dalton Trans.,
R. Mahiou and A. Gautier, Dalton Trans., 2009, 6894–
2009, 7308–7313.
6
902.
22 B. Bertrand, L. Stefan, M. Pirrotta, D. Monchaud, E. Bodio,
P. Richard, P. Le Gendre, E. Warmerdam, M. H. de Jager,
G. M. M. Groothuis, M. Picquet and A. Casini, Inorg.
Chem., 2014, 53, 2296–2303.
9
K. M. Hindi, M. J. Panzner, C. A. Tessier, C. L. Cannon and
W. J. Youngs, Chem. Rev., 2009, 109, 3859–3884.
0 S. Ray, R. Mohan, J. K. Singh, M. K. Samantaray,
1
M. M. Shaikh, D. Panda and P. Ghosh, J. Am. Chem. Soc., 23 S. Trapp and R. W. Horobin, Eur. Biophys. J. Biophys. Lett.,
007, 129, 15042–15053.
2005, 34, 959–966.
2
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.