N,N-Dialkylbenzimidazol-2-ylidene platinum complexes-effects of alkyl residues and ancillary cis-ligands on anticancer activity

Article abstract of DOI:10.1039/c8dt03360a

Eleven complexes of [(1,3-dialkylbenzimidazol-2-ylidene)L_nCl_3?n]Pt^(n?1)+, with L_n = DMSO (8), Ph₃P (9), (Ph₃P)₂ (10), and alkyl = Me (a), Et (b), Bu (c), octyl (d), were synthesised and tested for cellular accumulation, cytotoxicity, interference with the tumour cell cycle, and interaction with DNA. The delocalised lipophilic cationic bisphosphane complexes 10 were on average found to be more cytotoxic in MTT assays against a panel of seven cancer cell lines than the neutral DMSO and monophosphane complexes 8 and 9. The uptake of complexes 10, at least into HCT116 colon carcinoma cells, was also significantly greater than that of analogues 8 and 9. Their cytotoxicities did not differ significantly with the N-alkyl side chain length. The complexes that were most active, with sub-micromolar IC₅₀ (72 h) values against HCT116^wt cells, that is 8b, 9b, 10a-c, worked by a mode of action that was dependent on the functional p53, yet were still far more active than cisplatin in both of the HCT116^wt and HCT116^?/? variants. In detailed binding analyses 8c, 9c and 10a-c showed a lower affinity to DNA and different binding modes when compared to cisplatin, preferably forming mono-adducts with DNA and distorting it to a lower extent. Also, unlike cisplatin, they arrested the HCT116 cells of both variants predominantly in the G1 phase.

Full text of DOI:10.1039/c8dt03360a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
N,N-Dialkylbenzimidazol-2-ylidene platinum
complexes – effects of alkyl residues and ancillary
cis-ligands on anticancer activity†
Cite this: Dalton Trans., 2018, 47,
17367
b
Tobias Rehm,‡a Matthias Rothemund,‡a Alexander Bär,a Thomas Dietel,
Rhett Kempe,b Hana Kostrhunova, c Viktor Brabec, c Jana Kasparkova *c and
a
Rainer Schobert
*
Eleven complexes of [(1,3-dialkylbenzimidazol-2-ylidene)LnCl3−n]Pt(n−1)+, with Ln = DMSO (8), Ph3P (9),
(Ph3P)2 (10), and alkyl = Me (a), Et (b), Bu (c), octyl (d), were synthesised and tested for cellular accumu-
lation, cytotoxicity, interference with the tumour cell cycle, and interaction with DNA. The delocalised
lipophilic cationic bisphosphane complexes 10 were on average found to be more cytotoxic in MTT
assays against a panel of seven cancer cell lines than the neutral DMSO and monophosphane complexes
8 and 9. The uptake of complexes 10, at least into HCT116 colon carcinoma cells, was also significantly
greater than that of analogues 8 and 9. Their cytotoxicities did not differ significantly with the N-alkyl side
chain length. The complexes that were most active, with sub-micromolar IC50 (72 h) values against
HCT116wt cells, that is 8b, 9b, 10a–c, worked by a mode of action that was dependent on the functional
p53, yet were still far more active than cisplatin in both of the HCT116wt and HCT116−/− variants. In
detailed binding analyses 8c, 9c and 10a–c showed a lower affinity to DNA and different binding modes
when compared to cisplatin, preferably forming mono-adducts with DNA and distorting it to a lower extent.
Also, unlike cisplatin, they arrested the HCT116 cells of both variants predominantly in the G1 phase.
Received 16th August 2018,
Accepted 13th November 2018
DOI: 10.1039/c8dt03360a
rsc.li/dalton
in the case of cisplatin, the substitution of the ammine resi-
dues by mono- or bidentate amines and/or the exchange of the
Introduction
Since its approval in 1979, cisplatin (1, CDDP) has established chlorido ligands by chelating carboxylic acids. Since the turn
itself as the leading anti-cancer drug for several cancer entities. of the millennium, ever more diverse ligands with targeting
In nearly 40 years only six more approved platinum drugs have functions or bioactivity of their own have been attached to PtII
arisen from intensive scientific research. Although these drugs complexes and PtIV prodrugs.2
offer quite a few advantages over cisplatin, such as increased
With the advent of the stable N-heterocyclic carbenes
tumour specificity and reduced unwanted side effects, they (NHCs), first being made accessible by Arduengo et al.,3 a gen-
are still hampered by their high toxicity and adverse effects uinely new ligand motif became available that allowed the
in patients, as well as a tendency to elicit tumoral drug stereoelectronic fine-tuning of catalysts4 and the pharmaco-
resistance.1
logical optimisation of metallodrugs. Owing to their ease of
A common strategy to overcome these negative effects is the synthesis, their structural variability and their chemical stabi-
variation of the spectator and the leaving ligands, for example, lity, N-heterocyclic carbene (NHC) complexes of metals5 as
diverse as silver,6 gold,7 copper,8 ruthenium9 and platinum10
have become an ever-growing branch of medicinal chemistry.
aDepartment of Chemistry, University Bayreuth, Universitaetsstrasse 30,
N-Heterocyclic carbenes are often teamed up in square
95440 Bayreuth, Germany. E-mail: Rainer.Schobert@uni-bayreuth.de
bLehrstuhl fuer Anorganische Chemie II, University Bayreuth, Universitaetsstrasse 30,
planar PtII complexes with other ligands such as amines,10a,b
pyridines10c or pnictogen ligands10d,e (PPh3, AsPh3, SbPh3). In
95440 Bayreuth, Germany
cInstitute of Biophysics, Academy of Sciences of the Czech Republic, CZ-61265 Brno,
Czech Republic. E-mail: jana@ibp.cz
the past, we have concentrated on PtII complexes carrying
DMSO or phosphane ligands cis to benzylated imidazol-2-
ylidene ligands, for example 2–4.10d We have also developed a
protocol to introduce a second, optionally different, NHC
ligand in the cis-position to the first one as show in complex 5
(Fig. 1).11 We have now found that complexes with N-alkylated
†Electronic supplementary information (ESI) available: Ligand synthesis; NMR
spectra; X-ray structural data, values for cellular accumulation, cell cycle analysis
in HCT116 cells. CCDC 1857511–1857513. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c8dt03360a
‡These authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17367

View Article Online
Paper
Dalton Transactions
Fig. 1 Structures of cisplatin (1) and the NHC complexes 2–5.
Scheme 1 Synthesis of complexes 8a–d, 9a–d and 10a–c. (i) Ag2O,
CH2Cl2, rt; (ii) K2PtCl4, DMSO, 60 °C; (iii) PPh3, CH2Cl2; (iv) PPh3, CH2Cl2,
rt; 10c: CH3CN, 60 °C; 10d: unstable, decomposed upon attempted
isolation.
benzimidazoles rather than imidazole ligands show some
advantages, including
a better solubility combined with
chemical stability and the tendency to form easy to purify, and
easy to weigh out, crystalline complexes. The aim of this study
was to evaluate the influence of structural changes, for
example the alkyl chain length, overall size, lipophilicity, and
steric encumbrance of the benzimidazol-2-ylidene complexes
of PtII, on the effects of the complexes on cancer cells. To this
end, we synthesised the ligand precursors 6a–d and their
downstream complexes of types 8–10 (Scheme 1). By extending
the length of the N-alkyl chain from the methyl to the ethyl,
butyl and up to the octyl we gradually increased the size and
lipophilicity of the NHC ligands and their respective platinum
of the silver carbene complex 7d took several days, probably
owing to the bulkiness of its NHC ligand.
Transmetalation to the respective cis-[(NHC)(DMSO)PtIICl2]
complexes 8 was achieved by adding the silver complexes 7 to
K2PtCl4 in DMSO to give cis-Pt(DMSO)2Cl2 as a reactive inter-
mediate. The crude silver complex 7a of uncertain purity was
treated only with 0.9 equivalents of K2PtCl4 to prevent con-
tamination of the product complex 8a by leftover cis–Pt
(DMSO)2Cl2. After extraction and crystallisation all of the com-
plexes of 8 were obtained in good yield and a purity sufficient
for biochemical assessment.
complexes.
A comparison of the complex patterns cis-
[Cl2(NHC)(DMSO)]PtII, cis-[Cl2(NHC)(PPh3)]PtII and trans-[Cl
(NHC)(PPh3)2]Pt+Cl−, analogous to 2–4, should allow a predic-
tion of the influence of the lipophilicity, steric encumbrance,
and charge on the biological properties of similar NHC-PtII
anticancer drugs. Eleven such NHC-platinum(II) complexes
with benzimidazol-2-ylidene ligands of varying size and lipo-
philicity were prepared and studied. Complex 10d was inac-
cessible as it decomposed upon attempted isolation.
For the synthesis of the phosphane complexes of types 9
and 10 we intended to apply the same protocols as for the
preparation of the imidazol-2-ylidene analogues 3 and 4.
Complex 3 had been obtained by substitution of the labile
DMSO ligand of complex 2 with 1.1 equivalents of PPh3 in
CH2Cl2 at room temperature within 90 min. The conversion of
complex 2 to the cationic bisphosphane complex 4 required
stirring with five equivalents of PPh3 for 30 min. However, the
benzimidazol-2-ylidene complexes 9 and 10 were not accessi-
ble in this way. When treated with 1.1 equivalents of PPh3 only
50% of the DMSO complex 8a, bearing a 1,3-dimethyl-
benzimidazol-2-ylidene ligand, reacted to afford the exclusively
Results and discussion
Synthesis and characterisation
Benzimidazolium salts with methyl, ethyl, butyl, and octyl resi- bisphosphane complex 10a, but none of the monophosphane
dues were obtained by treating benzimidazole with K2CO3 and complex 9a, which is obviously more reactive than 8a. To
either iodomethane, iodoethane, 1-bromobutane, or 1-bromo- prepare complex 9a in an acceptable yield, a solution of PPh3
octane, respectively. They were converted to their chlorides was added slowly to a highly diluted solution of 8a. The
6a–d to prevent anion exchange during the following steps. product was then separated and purified using column chrom-
Treating them with Ag2O in CH2Cl2 for at least 24 h afforded atography and crystallisation. When the sterically more hin-
the NHC silver(I) complexes 7a–d after filtration and crystallisa- dered N,N-diethyl substituted analogue 8b was reacted over-
tion. As complex 7a was poorly soluble it could not be separ- night with five equivalents of PPh3, the monophosphane
ated from the silver residues and was therefore converted to complex 9b was isolated in good yield besides only trace
the platinum complex 8a without purification. The formation amounts of the cationic bisphosphane complex 10b. To gain
17368 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
the latter in a moderate yield 8b was stirred with 10 equiva- Fig. 2. Repeated attempts to isolate the cationic product from
lents of PPh3 for a full five days, followed by careful crystallisa- such solutions, with only traces of 9d and PPh3 left, led to
tion. Unsurprisingly, the sterically even more hindered DMSO equilibrated mixtures again, or even pure 9d in significant
complex 8c, bearing
a
1,3-dibutylbenzimidazol-2-ylidene amounts. This indicates that 10d is stable in solution only in
ligand, when reacted with 10 equivalents of PPh3 for several the presence of an excess of PPh3 and decomposes upon iso-
days afforded the monophosphane complex 9c exclusively. The lation. It was therefore not possible to purify this compound
cationic bisphosphane complex 10c could be prepared in for biological studies.
acceptable yield only by heating the DMSO complex 8c with
The new PtII complexes 8, 9 and 10 were characterised
20equivalents of PPh3 in acetonitrile at 80 °C for four days. using 1H, 13C, 195Pt and 31P nuclear magnetic resonance
Similarly, the DMSO complex 8d, bearing a 1,3-dioctylbenzi- (NMR) spectroscopy, as well as mass spectrometry and elemen-
midazol-2-ylidene ligand, reacted with five equivalents of PPh3 tal analysis. Single crystal X-ray structure analyses were per-
in CH2Cl2 at room temperature for 16 h to give monophosphane formed for complexes 8b, 9c and 10a (Fig. 3).
complex 9d. A second PPh3 was incorporated only upon
The completion of the formation of the DMSO complexes 8
heating with a large excess of PPh3 in acetonitrile at 80 °C. can be seen from the equivalent benzimidazole and DMSO
Attempts to isolate the bisphosphane complex 10d always gave signals in the 1H NMR spectra. For 8b–d the N–CH2 protons of
mixtures of 10d, 9d and PPh3. 31P NMR spectra revealed that the side chains were inequivalent and gave rise to two sets of
in solution, with an excess of phosphane, the signals for 9d signals. As seen before with the benzylated imidazole
and PPh3 at 9.1 ppm and −5.5 ppm, respectively, would disap- ligands,10d this confirms the cis configuration of the NHC and
pear over time to generate more 10d at 18.9 ppm, as shown in DMSO ligands as well as a perpendicular orientation of the
benzimidazole relative to the plane spanned by the
PtCl2(DMSO) fragment. This effect recurs for 9b–d, while the
spectra of the symmetric complexes 10 show no inequivalent
protons.
To monitor the formation of complexes 9 and 10 and to
also verify their conformation, 31P NMR spectroscopy is a con-
venient technique. The neutral complexes 9 exhibit a signal
between 8.83 and 9.19 ppm with coupling constants ranging
from 3861 to 3901 Hz. The cationic complexes 10 with two
equivalent PPh3 moieties show a signal between 17.7 and
18.9 ppm, coupling with a JP–Pt = 2478 Hz to 2486 Hz typical of
the trans-configured phosphanes.10d,e
The same Pt–P couplings are present in the 195Pt NMR
spectra. Complexes 8a–d give a singlet around −3554 to
Fig. 2 31P NMR spectra of reactions over time between PPh3 and 9d to
give 10d, and for the decomposition products obtained upon attempts
to isolate it from the “t = 14 d” sample.
−3566 ppm, complexes 9 show the expected doublets at −4017
to −4018 ppm and 10 shows triplets at −4378 to −4383 ppm.
With JPt–P = 3871–3895 Hz for 9 and 2486–2490 Hz for 10 the
Fig. 3 Molecular structures of complexes 8b, 9c and 10a as thermal ellipsoid representations at the 50% probability level (H atoms omitted).
Selected bond lengths [Å] and angles [°]: 8b: Pt1–Cl1 2.3447(19), Pt1–Cl2 2.332(2), Pt1–C1 1.983(7), Pt1–S1 2.202(2), Cl1–Pt1–Cl2 90.71(7), Cl1–Pt1–
S1 90.35(7), Cl1–Pt1–C1 177.6(2), Cl2–Pt1–C1 86.9(2), Cl2–Pt1–S1 178.84(7), C1–Pt1–S1 92.1(2), N1–C1–N2 107.7(6); 9c: Pt1–C1 1.980(2), Pt1–P1
2.2345(8), Pt1–Cl1 2.3603(8), Pt1–Cl2 2.3480(8), C1–Pt1–P1 95.34(7), C1–Pt1–Cl1 86.76(7), C1–Pt1–Cl2 175.80(6), P1–Pt1–Cl1 175.686(19), P1–Pt1–
Cl2 88.09(4), Cl1–Pt1–Cl2 89.65(3), N1–C1–N2 106.70(19); 10a: Pt1–Cl1 2.349(3), Pt1–C1 1.959(10), Pt1–P1 2.309(3), Pt1–P2 2.311(3), C1–Pt1–Cl1
179.4(4), C1–Pt1–P1 92.2(3), C1–Pt1–P2 90.7(3), P1–Pt1–P2 176.87(10), P1–Pt1–Cl1 87.77(10), P2–Pt1–Cl1 89.33(9), N1–C1–N2 105.3(9).
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17369

View Article Online
Paper
Dalton Transactions
coupling constants agree with those obtained from the 31P lines, while the addition of a second PPh3 ligand, as seen in
NMR spectra and with the literature values of similar complex- the cationic complexes 10, gave rise to low nanomolar IC50
es.10d,e In the 13C NMR spectra the carbene signals are of values against most cell lines. The efficacy of complexes 10 was
special interest. They were found at around 154 ppm for com- superior to that of CDDP in all cell lines except for the resist-
plexes 8, between 160.6 ppm and 162.0 ppm for 9, and ant Kb-V1Vbl cervix carcinoma cell line.
between 158.0 and 159.9 ppm for the cationic complexes 10.
There was no clear trend for the effect of the N-alkyl side
chain length on the cytotoxicity of the individual complex
types 8, 9, and 10 over all of the tested cell lines. The neutral
Crystallography
Crystals suitable for X-ray diffraction analyses were grown by N-methyl derivatives 8a and 9a were inactive (IC50 > 100 µM) in
slow infusion of hexane or diethyl ether into saturated solu- all tested cell lines, while the lipophilic delocalised cation 10a
tions of 8b, 9c or 10a in CH2Cl2 kept at 4 °C. Fig. 3 shows their was highly active with low nanomolar IC50 values in all but the
molecular structures. The distances between the platinum and cervix carcinoma cell line. The N-ethyl substituted complex 8b
the carbene carbon atoms were in the range of 1.959 to was the most active DMSO complex in all of the cell lines bar
1.983 Å. The Pt–Cl distances lay between 2.347 to 2.349 Å for the 518A2 melanoma in which it was virtually inactive. Its high
those trans to the NHC ligand, while the cis coordinated varied cytotoxicity in the slow-growing colon carcinoma cell lines
from 2.332 Å in the DMSO-complex 8b to 2.360 Å in 9c trans to DLD-1 and HCT116wt, as well as in the re-sensitised cervix car-
a phosphine. In cationic 10a, Pt–P distances were both around cinoma cell line Kb-V1Vbl+Vpm, is worthy of note. Elongation of
2.31 Å, while in neutral 9c, this was 2.234 Å and the Pt–S bond the side chain as in 8c and 8d had no distinct effect.
in 8b was even shorter at 2.202 Å. The bond angles between
Amongst the generally bigger and more lipophilic mono-
the carbene and the cis-phosphine or DMSO ligand were phosphine complexes the N-ethyl substituted 9b was roughly
slightly above 90° with a maximum of 95.3° for 9c.
as active as its N-butyl analogue 9c while complex 9d, the
biggest and most lipophilic, was distinctly less active. The
delocalised lipophilic cations 10 were less affected by changes
Cytotoxicity
The new benzimidazol-2-ylidene PtII complexes 8a–d, 9a–d, at their NHC ligands and showed similar IC50 values in the
10a–c, and CDDP as a reference, were evaluated using a MTT- nanomolar range, or the low micromolar range in the case of
assays for their in vitro cytotoxicity against a panel of seven complex 10c.
cancer cell lines of various entities, including a p53 knock-out
Judging by their IC50 values, the complexes 8b–d, 9b–d and
mutant (HCT116−/−) and the two multi-drug resistant (mdr) 10a–c were as active against the mdr breast cancer cell line
cell lines Kb–V1Vbl and MCF-7Topo. The complexes showed a MCF–7Topo as against all other cell lines, suggesting that they
structure-dependent cytotoxicity, with IC50 values ranging from are not recognised as substrates by the breast cancer resistance
inactive (IC50 > 100 µM) to low nanomolar (Table 1). A trend protein (BCRP) which is overexpressed in this cell line.13a In
emerged with respect to the three complex types 8, 9, and 10, contrast, all of these complexes were conspicuously less
reminiscent of that previously observed for the imidazol-2- effective against the mdr cervix carcinoma cell line Kb–V1Vbl
ylidene complexes 2–4.10d Substitution of the DMSO ligand in which overexpresses P-glycoprotein 1 (P-gp1), presumably
complexes 8 by a PPh3 ligand led to a slight increase of the because they are substrates of this efflux transporter.13b,c To
overall cytotoxicity of the resulting complexes 9 in some cell corroborate this assumption, the IC50 values were determined
Table 1 Mean SD of IC50 (72 h) values [µM] of complexes 8, 9, 10a–c and CDDP in MTT assays against human cancer cell linesa as calculated
from four independent measurements
IC50 (72 h) [µM]
Compounds
518A2a
HT-29a
DLD-1a
HCT116wt a
HCT116−/− a
MCF-7Topo a
Kb-V1Vbl a
Kb-V1Vbl+Vpm a
rp53 b
8a
8b
8c
8d
>100
>50
7.6 0.2
3.1 0.1
>100
>100
>100
>100
n.d.
n.d.
n.d.
n.d.
—
8.6 1.7
29.4 0.9
11.6 0.5
>100
0.24 0.03
12.7 0.9
11.0 1.0
>100
0.12 0.01
8.1 0.7
3.9 0.3
>100
5.9 0.5
6.0 0.2
9.9 0.7
n.d.
6.5 1.6
10.1 0.3
8.1 0.4
n.d.
25.2 0.6
21.2 0.5
28.3 6.8
n.d.
0.28 0.06
0.25 0.09
4.5 0.9
n.d.
−0.979
0.356
−0.627
—
9a
9b
9c
9d
10a
10b
10c
CDDP
1.8 0.2
2.5 0.1
11.0 0.3
0.21 0.03
0.29 0.04
0.24 0.03
2.6 0.7
1.9 0.4
1.3 0.0
11.0 0.5
0.17 0.01
0.16 0.04
0.08 0.01
8.3 0.5
1.5 0.1
0.61 0,06
9.4 0.7
0.30 0.06
0.45 0.14
2.6 0.1
5.5 0.2
0.91 0.09
2.6 0.8
12.3 1.2
0.15 0.01
0.20 0.03
0.15 0.01
5.7 0.3
3.7 0.3
2.4 0.4
12.3 0.5
0.28 0.03
0.50 0.04
2.1 0.2
9.2 0.5
3.9 0.3
1.5 0.1
14.5 3.1
0.10 0.01
0.29 0.05
1.1 0.1
13.5 1.7
20.4 1.2
11.5 0.3
14.1 2.8
18.5 0.9
4.6 0.2
2.6 0.2
0.13 0.01
5.6 0.3
1.6 0.3
4.5 0.9
1.10 0.2
0.12 0.02
0.13 0.03
0.23 0.08
−0.755
0.058
−0.002
−0.464
−0.599
−0.928
−0.380
a 518A2 – melanoma, HT-29 – colon adenocarcinoma, DLD-1 – Dukes type C colorectal adenocarcinoma, HCT116wt – colon carcinoma (wildtype);
HCT116−/− – colon carcinoma (p53 knock-out mutant); MCF-7Topo – mamma carcinoma, Kb-V1Vbl – cervix carcinoma. b p53 dependency of IC50
values in HCT116 cells, indicated by rp53 = (IC50wt/IC50−/−) − 1.
17370 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
once more in the presence of the competitive P-gp1 inhibitor
verapamil and found to be significantly lower when compared
to those obtained without added verapamil (Table 1, second
column from the right).
p53 dependency
To investigate the dependency of the cytotoxic activity on p53,
which represents one activator of the apoptotic cascade after
DNA damage and prolonged cell cycle arrest, and which is nor-
mally activated upon treatment with CDDP,14,15 the IC50 values
of 8b–d, 9b–d, 10a–c and CDDP were determined in the
HCT116−/− p53 knock-out mutant cell line and the p53 depen-
p53(= (IC50wt/IC50
) − 1) was calculated
−/−
Fig. 5 Intracellular platinum levels in HCT116wt colon carcinoma cells
after 5 h of incubation with 8 µM of Pt(II) complexes CDDP, 8a–c, 9a–d
and 10a–c, measured via ICP-MS. Mean SD from three independent
experiments. Exact values can be found in ESI Table S2.†
dency factor
r
(Table 1). In Fig. 4 the rp53 values are depicted as bars to illus-
trate the differences in the IC50 values in the wildtype (wt) and
the p53 knock-out mutant (−/−). The longer the bar the
greater the effect that the deletion of p53 has on the cytotoxic
activity of each complex. Bars in the negative range indicate a
lower cytotoxicity in the p53−/− mutant in comparison to the
wildtype cells, bars in the positive range indicate a higher cyto-
toxicity. Owing to the inactivity of 8a and 9a against the wild-
type cell line these complexes were not tested in the p53 nega-
tive cell line. Except for 9d, all tested compounds seemed to be
dependent on p53 to various degrees. The cationic complexes
10a–c showed a correlation between the length of the N-alkyl
residues and the difference between the toxicity in the wildtype
cells and the p53−/− cells.
length, variation of the secondary ligands (DMSO, PPh3,
(PPh3)2), and a potential positive charge. Fig. 5 shows the
amount of Pt (pmol per 106 cells) taken up by the cells after
5 h of incubation with the compounds (cf. ESI† for values SD).
Complexes 8a–c, carrying DMSO as an ancillary ligand, are
not taken up much better than CDDP. The increased toxicity of
8b for example, over that of CDDP or 8a is owed to mechanistic
reasons rather than the uptake rates. Within the group of com-
plexes 9, the uptake rises with the growing length of the
N-alkyl chain from methyl (9a) to butyl (9c). However, the
N-octyl substituted complex 9d is taken up to a far lesser
extent, ranging between that of CDDP and 9a. However, when
the cytotoxicities of 9b–d measured against HCT116wt were
adjusted to their uptake rates the resulting relative ‘intrinsic’
cytotoxicities were found to be quite similar.
A comparison of complexes 8, 9 and the charged complex
10, shows that the exchange of the secondary ligands (DMSO
or Cl) for one or two phosphanes vastly promotes the cellular
uptake of the platinum complexes, independent of the chain
length. Almost six times as much of the bisphosphino
complex 10c was taken up as that of 8a–c and about 12 times
as much as the of CDDP. Consequently, the ‘intrinsic’ cytotoxi-
cities against HCT116wt differed less markedly than the actual
measured ones. The upshot is that we found a weak correlation
between the uptake and the N-alkyl side chain length for com-
plexes a–d, yet no decisive influence was found for the side
chains ranging from ethyl to octyl on the cytotoxicity.
Cellular uptake
In order to check whether the observed differences in the cyto-
toxicities of the new complexes are a consequence of different
uptake rates and intracellular concentrations, the latter were
measured via inductively coupled plasma mass spectrometry
(ICP-MS) in HCT116wt cells treated with 8a–c, 9a–d, 10a–c and
CDDP. These complexes cover the whole spectrum of chain
DNA interaction
EMSA and ethidium bromide assays. One of the main
targets of CDDP is the nuclear DNA. By intra- and inter-strand
crosslinking, CDDP drastically changes the DNA morphology,
Fig. 4 Comparison of the IC50 values (72 h) of 8c–d, 9b–d, 10a–c, and and essential cellular processes, such as DNA replication or
CDDP against the HCT116 colon carcinoma cells with wildtype p53 (wt)
protein expression are hindered, and ultimately apoptotic
and the p53 knock out mutant (−/−) via the dependency factor rp53
mechanisms are activated.14c We examined the potential DNA
which is defined as (IC50wt/IC50−/−) − 1. The further rp53 deviates from
interaction of complexes 8, 9, 10 and CDDP using ethidium
bromide saturation assays with salmon sperm DNA (Fig. 6 top)
and electrophoretic mobility shift assays (EMSA) with circular
zero, the wider apart the two IC50 values are set. Negative rp53 values
indicate a lower cytotoxicity of the compound in the knock out mutant,
positive values indicate a higher cytotoxicity.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17371

View Article Online
Paper
Dalton Transactions
Fig. 6 Top: Ethidium bromide – DNA adduct fluorescence after 2 h of
incubation with complexes 8, 9, 10 or CDDP, relative to a vehicle control
(set to 100%). Mean SD calculated from 3 independent measurements.
Bottom: EMSA of circular pBR322 plasmid DNA after 24 h of incubation
with 0, 5, 25 and 50 µM of 8c, 9c or 10c. Staining of DNA bands was per-
formed using ethidium bromide.
Fig. 7 (A) Quantification of Pt from 8c, 9c and 10a–c bound to DNA in
0.01 M NaClO4 at 37 °C after 2 or 24 h. Samples were dialysed against
water (columns 2 h and 24 h) or against 0.1 M NaCl (columns 2 h/Na+ or
24 h/Na+). The platinum content associated with the DNA was deter-
mined using FAAS. (B) Progress of DNA platination by 8c, 9c, and 10c.
Samples, withdrawn after indicated time intervals were exhaustively dia-
lysed against 0.1 M NaCl and analysed using FAAS.
pBR322 plasmid DNA (Fig. 6 bottom). Complexes 8a–c neither
reduced the ethidium bromide (EtBr) fluorescence after incu-
bation with salmon sperm DNA, nor caused any band shifts or
aggregations of the plasmid DNA in EMSA. This coincides ment of the Pt content using flame atomic absorption spec-
nicely with the lack of DNA affinity that we found recently for troscopy (FAAS) after 2 and 24 h. As indicated in Fig. 7A, there
the related complexes 2.10d Although 9a, 9b and 9d showed no was a significant difference in the DNA binding between the
impact in the EtBr saturation assay, complex 9c reduced the neutral complexes 8c and 9c and the cationic compounds
EtBr fluorescence to a similar extent as the charged complexes 10a–c. After a short incubation time, approximately 40% of 8c
10. Despite the negative result from the EtBr saturation assay, and 9c were bound to the DNA, while ca. 80% were associated
9b (ethyl) induced, like 9c (butyl), a band shift in the EMSA at with it after 24 h. The amount of platinum bound to the DNA
a concentration of 25 µM. However, complex 9d, carrying the did not change after dialysis with a high salt buffer. In con-
longest N-alkyl side chain, showed virtually no effect in the trast, complexes 10a–c associated much faster with DNA, but
EMSA. The positive charge of complexes 10 likely facilitates when the treated DNA samples were dialysed after 2 h against
their interaction with the negatively charged phosphate back- a buffer of high ionic strength, the amount of Pt that stuck to
bone of the DNA,16 resulting in a reduction of the EtBr fluo- the DNA was much lower (ca. 5%). These results indicate that
rescence to about 15% of the vehicle control value, and a band complexes 10a–c associate to DNA in a fast, though initially
shift in the EMSA as well as the formation of DNA aggregates, mainly electrostatic, manner. Under the conditions of a high
which are too bulky to leave the pockets of the agarose gel Na+ concentration, these electrostatic attractions are disrupted
during electrophoresis.
by saturation of the negatively charged phosphate backbone of
Kinetics and DNA base preferences. As neither the EMSA the DNA, and only the Pt coordinatively bound to the DNA
nor the ethidium bromide saturation assays gave detailed remains. The coordinative binding of 10a–c is much slower
information on the DNA–complex interaction, more sophisti- than that of 8c and 9c, this is probably due to the steric hin-
cated biophysical methods were applied, with a focus on com- drance of the two bulky PPh3 groups of 10 (Fig. 7B). Next, we
plexes 8c, 9c and 10a–c which cover the whole spectrum of investigated the preferences of complexes 8c, 9c, and 10a–c for
ancillary ligands (DMSO, PPh3, (PPh3)2) while sharing a butyl binding to particular nucleotide sequences by exposing them
chain, but also allow us to gauge the influence of the side to various synthetic single-stranded homopolydeoxyribo-
chain length.
nucleotides. All of the tested complexes bound readily to single-
The speed and degree of platination of double-stranded calf stranded poly(dG) and, to a lesser extent, to poly (dA), but vir-
thymus (ct) DNA was investigated using its incubation with tually not at all to the single-stranded poly(dT) or poly(dC)
defined concentrations of the respective complexes, dialysis (cf. ESI† for exact values), in keeping with the base sequence
against water or high salt buffer (0.1 M NaCl), and measure- preferences of cisplatin.17
17372 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
DNA stability. The capability of 8c, 9c, 10a–c to distort and than before (cf. Fig. 6), the degree of supercoiling of the nega-
destabilise dsDNA was assessed by measuring the absorbance tively supercoiled pSP73 plasmid DNA was analysed by agarose
of the platinated ct DNA at 260 nm as a function of the tem- gel electrophoresis (cf. ESI Fig. S43†).19 Any unwinding and
perature. The Pt/nucleotide ratio was set to rb = 0.03 and the reduction of the supercoils in a closed circular plasmid affects
samples were dialysed against 0.1 M NaCl and, subsequently, its electrophoretic mobility. The decrease of the mobility
against 0.01 M NaClO4 to get rid of any unbound complex during electrophoresis allows the quantification of the mean
molecules. The effect of the resulting DNA platination on its unwinding angle θ per adduct. It can be calculated as θ =
melting temperature (Tm) was measured in samples with low −18σ/rb(c), in which σ is the superhelical density and rb(c) is
(0.01 M) and high (0.1 M) concentrations of Na+. In contrast to the value of rb at which the supercoiled and nicked forms
CDDP, the complexes 8c, 9c, and 10a–c had a slightly comigrate.19 Under the conditions used in these experiments,
increased Tm at low Na+ concentrations (cf. ESI Fig. S41,† left). the value for σ of the plasmid DNA was calculated to be −0.029
At high Na+ concentrations the Tm decreased as with CDDP, on the basis of the data for CDDP, for which the rb(c) was
but not to the same degree (cf. ESI Fig. S41,† right). The obser- determined in this study and θ = 13° was calculated.19 Using
vation that the Tm values of DNA modified by 8c, 9c, and 10a–c this approach, the DNA unwinding angles were determined to
decreased at a higher salt concentration (0.1 M) is consistent be 6.2 0.5°, 6.1 0.3°, 5.6° 0.5°, 6 1°, and 6.6 1.0° for
with the presence of conformational alterations induced by 8c, 9c, 10a, 10b, and 10c, respectively, which are thus signifi-
these complexes that destabilise the duplex. At higher salt con- cantly smaller than that found for cisplatin (13°), but very
centrations (0.1 M), the negative DNA phosphate groups are similar to those of the monofunctional complexes [Pt(NH3)3Cl]
already efficiently neutralised by the cations present in the Cl and [Pt (dien)Cl]Cl, for which unwinding angles of 6° have
medium, so that any stabilising electrostatic contribution by been measured.19 Interestingly, none of the complexes tested
the coordinated positively charged platinum fragments is less in this work increased the mobility of the relaxed form (oc) as
pronounced than in a medium of a lower salt concentration. CDDP does. It has been shown that the bifunctional binding
However, the destabilisation caused by 8c, 9c, and 10a–c was of CDDP and other platinum complexes to DNA shortens and
less prominent then that induced by cisplatin, indicating a condenses the DNA helix.20 The observation that the binding
lower extent of DNA distortion. In contrast, at low salt concen- of 8c, 9c, and 10a–c does not affect the mobility of oc DNA
trations, in which the negative charges of the DNA phosphate suggests that these complexes predominantly form DNA mono-
groups are not sufficiently neutralised by the Na+ counterions, adducts, even though 8c and 9c contain two leaving chloride
the electrostatic stabilising effects of 8c, 9c, and 10a–c override ligands.
their destabilising effects stemming from conformational dis-
DNA interstrand crosslinking. Apart from the stabilising
tortions. These results suggest that the Pt(II)-benzimidazole effect of the coordinated positively charged platinum complex
complexes induce conformational distortions in dsDNA that fragments, and the destabilising conformational distortions,
differ from those induced by CDDP.
crosslinks are a third factor that significantly affect the
DNA distortion and unwinding. The DNA morphology altera- thermal stability of DNA. It has been shown that interstrand
tions upon binding of the investigated Pt(II) complexes were crosslinks (ICLs) in DNA markedly increase its melting tem-
further assessed according to their effects on the fluorescence perature by preventing the dissociation of the strands.21 For
of DNA-terbium ion (Tb3+) adducts and on the unwinding of complexes 10a–c, which contain only one leaving chloride
pSP73 double helical plasmid DNA. Tb3+ fluorescence has ligand, the formation of ICLs in double helical DNA is un-
been used to investigate local perturbations induced in likely, in contrast to complexes 8c and 9c which feature two
double-helical DNA by various physical or chemical agents replaceable chlorido ligands. We tested the ability of com-
including platinum(II) complexes. The fluorescence of Tb3+ plexes 8c, 9c, and 10a–c, as well as CDDP, to crosslink
ions is strongly enhanced when they are bound to unplatinated pSP73KB DNA which had been linearized and 32P-labeled. The
guanine (G) residues in distorted DNA regions.18 Ct DNA was electrophoretic autoradiograms shown in Fig. 8 reveal that
treated with the complexes at rb values of 0.03. Unbound plati- CDDP, even at very low concentrations rb, gave rise to distinct
num was removed by dialysis as described in the materials interstrand-crosslinked DNA bands which migrate more slowly
and methods. The platinated DNA samples were treated with through the denaturing agarose gel than single stranded (ss)
TbCl3, and subsequently, the fluorescence of the Tb3+-DNA DNA. In contrast, complexes 8c and 9c caused comparably
adducts was measured (cf. ESI Fig. S42†). The treatment of fainter ICL DNA bands, even when applied at rb values ten
DNA with complexes 8c, 9c, and 10a–c resulted in a slight times of those for CDDP. Although the frequency of the ICLs
enhancement of the terbium fluorescence, which was, caused by CDDP was 6%, it was far less than 1% for complexes
however, less pronounced than that induced by CDDP. These 8c and 9c. With complex 10c, as with 10a and 10b (not
results confirm that DNA adducts of these Pt(II)-benzimidazole shown), no ICL DNA bands were detected. It should be noted,
complexes give rise to local conformational alterations in though, that the absence of DNA crosslinking is not necess-
double-stranded DNA which are less distorting than those of arily an indication of a lower biological activity. It has been
CDDP.
shown that DNA adducts of monofunctional anticancer
To determine the effect of the DNA adducts of 8c, 9c, and Pt(II) complexes, such as pyriplatin, phenanthriplatin, or cis-
10a–c on the unwinding of double-helical DNA in more detail {Pt(NH3)2[3,5-dichloro-N-(3-chloro-4-(quinolin-6-yl-oxy)phenyl)-
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17373

View Article Online
Paper
Dalton Transactions
Fig. 9 Effects of 8c (10 µM), 9c (500 nM), 10c (500 nM) and CDDP
(10 µM) on the progression of the cell cycle of HCT116wt colon carci-
noma cells after 24 h of treatment. The bars represent the percentages
of cells in each phase of the cell cycle (G1, S and G2/M) and dead cells
(sub-G1). Analysis was performed via propidium iodide staining and flow
cytometry, values represent mean SDs of three experiments.
Fig. 8 Autoradiograms of a denaturing agarose gel of linear pSP73 DNA
modified by CDDP or complexes 8c, 9c, 10c. Interstrand crosslinked
DNA (upper bands, ICLs) migrates more slowly during electrophoresis
than single-stranded DNA (bottom bands, ss). Top: CDDP rb = 0, 0.001,
0.0015, 0.002; 8c rb = 0, 0.01, 0.015, 0.02; Bottom: 9c, 10c rb = 0, 0.01,
0.015, 0.02.
lines. With the exception of 9d, the uptake rates of all three
complex types into the HCT116 cells increased with the length
of their N-alkyl substituents. However, this effect was not mir-
rored by the course of their cytotoxicities against most of the
tested cancer cell lines which were somewhat erratic. Neutral
N-methyl substituted complexes 8a and 9a were virtually in-
active against all tested cancer cell lines, while the N-ethyl sub-
stituted complex 8b stood out as being specifically efficacious
against slow-growing colon carcinoma cell lines, where it
reached IC50 values like the cationic complexes 10. The com-
plexes that were most active with sub-micromolar IC50 (72 h)
values against HCT116wt cells were 8b, 9b, 10a–c, which
worked by a mode of action that was dependent on the func-
tional p53, which was evident from their reduced activity
against the HCT116−/− cancer cells. They were also far more
active than cisplatin in both of the HCT116 variants. In a cell-
free context, DNA was a major target for all of the new benzimi-
dazol-2-ylidene complexes, although their molecular inter-
actions differed considerably. Neutral complexes 8 and 9
appear to bind swiftly to DNA at G-residues in a coordinative
manner, whereas the cationic complexes 10 quickly pre-associ-
ate with DNA by electrostatic attraction, which is only slowly
replaced by coordinative binding. The DNA adducts formed
eventually in either case are characterised by monofunctional
platination, causing lesions which distort dsDNA to a much
lesser extent than those caused by cisplatin. This might go
hand in hand with reduced side effects in animals or humans.
Also, unlike cisplatin, complexes 10 and 9c arrested the
HCT116 cells of both variants in the G1 phase.
2-hydroxybenzamide]Cl}NO3,
bearing
bulkier
aromatic
ligands,22 effectively inhibit the RNA-polymerisation that is
considered to be crucial for the biological activity of Pt com-
plexes.23 Moreover, a less distinct DNA distortion might help
to hide the DNA lesions from the repair machinery.24
Cell cycle analysis
The binding of small molecules to the DNA and the resulting
damages affect various cellular DNA-dependent processes,
such as protein expression, DNA replication and mitosis.
Changes in the DNA integrity often lead to an arrest of the cell
cycle progression of the cells. Key players during the cell cycle
arrest and the activation of the apoptotic cascade are the
“guardian of the genome” p53 and the CDK1 inhibitor p21, a
major target of activated p53.15 Using propidium iodide stain-
ing and flow cytometry, we tested the N-butyl substituted com-
plexes 8c, 9c, and 10c for their effect on the cell cycle pro-
gression of HCT116wt cells (Fig. 9) and p53 knock-out mutant
HCT116−/− cells (cf. ESI, Fig. S44†). No distinct changes in the
cell cycle progression were observed upon treatment of the
wildtype cells with the DMSO complex 8c, while in the p53
knock-out mutant cells treatment led to a G2/M phase arrest.
In contrast, both phosphane complexes 9c and 10c at a con-
centration as low as 500 nM induced a strong G1 phase arrest
in the wildtype and the mutant cell lines after 24 h, indicating
a mode of action that differs from that of CDDP, which caused
an arrest in the S phase.
Upshot: SAR and mode of action
Conclusions
Some structure-uptake and structure–activity relations
emerged from our studies. The uptake of the delocalised lipo- The replacement of imidazol-2-ylidene by benzimidazol-2-
philic cationic complexes 10 by HCT116 colon carcinoma cells ylidene ligands in antitumoral platinum(II) complexes with
was conspicuously greater than that of the neutral analogues 8 otherwise identical ancillary and leaving ligands changes their
and 9. The complexes 10 were also—on average—more cyto- chemistry (synthesis, stability, and solubility) more than their
toxic in the MTT assays against the panel of seven cancer cell biological profile (DNA interaction, cancer cell cycle inter-
17374 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
ference, dependency of the cytotoxicity on the ancillary ligands CH3), 4.57 (4H, q, J = 7.2 Hz, CH2), 7.44–7.50 (2H, m, BI),
and on charge). However, their selectivity and individual 7.83–7.89 (2H, m, BI).
efficacies against particular cancer entities may differ signifi-
Chlorido(1,3-di-n-butylbenzimidazol-2-ylidene) silver(I) (7c). 6c
cantly. What makes the new N,N-dialkylbenzimidazol-2-ylidene (320 mg, 1.40 mmol), Ag2O (159 mg, 0.70 mmol, 0.5 eq.),
platinum complexes interesting as drug candidates is their CH2Cl2 (35 mL), 24 h; yield: 407 mg (1.09 mmol, 78%); amber
highly structure-dependent cytotoxicity against cancer cells, in solid; 1H NMR (500 MHz, CDCl3): δ 0.97 (6H, t, J = 7.3 Hz,
combination with reduced morphology changes to DNA, which CH3), 1.48–1.39 (4H, m, CH3–CH2), 2.01–1.90 (4H, m, NCH2–
bode well for the management of unwanted side effects in CH2), 4.54 (4H, t, J = 7.2 Hz, N–CH2), 7.41 (2H, dd, J = 6.1,
patients when compared to cisplatin.
3.0 Hz, BI), 7.53 (2H, dd, J = 6.0, 3.1 Hz, BI).
To achieve a rational structure-based optimisation of their
Chlorido(1,3-dioctylbenzimidazol-2-ylidene) silver(I) (7d). 6d
drug-like properties, in subsequent studies we need to clarify (240 mg, 633 µmol), Ag2O (147 mg, 633 µmol, 1 eq.), CH2Cl2
their uptake pathways and their intracellular distribution, and (15 mL), 3 d; yield: 238 mg (489 µmol, 77%); beige solid; 1H
in particular the potential contribution of antimitochondrial NMR (500 MHz, CDCl3): δ 0.84 (6H, t, J = 6.9 Hz, CH3),
effects by the cationic complexes 10.
1.19–1.29 (12H, m, CH2), 1.30–1.37 (4H, m, NCCCCH2),
1.37–1.44 (4H, m, NCCCH2), 1.98 (4H, quin, J = 7.4 Hz NCCH2),
4.53 (4H, t, J = 6.5 Hz, NCH2), 7.43 (2H, dd, J = 6.1, 3.2 Hz, BI),
7.53 (2H, dd, J = 6.1, 3.2 Hz, BI).
Experimental
Synthesis
of
cis-[Cl2(DMSO)(1,3-dialkylbenzimidazol-2-
Materials and methods
ylidene)]PtII complexes 8. A solution of silver complex 7 in
DMSO was treated with K2PtCl4 (1 eq.) and heated at 60 °C in
the dark for 20 h. After addition of CH2Cl2 the precipitating
AgCl was filtered off, the filtrate was washed with H2O and
brine, and dried over Na2SO4. Residual DMSO was removed
in vacuo and by crystallising the product from CH2Cl2 with
Et2O or hexane.
All chemicals and reagents were purchased from Sigma
Aldrich, Alfa Aesar, ChemPur or ABCR and were used without
further purification. Synthetic double-helical polydeoxyribo-
nucleotides poly (dA), poly (dT), poly (dC), and poly (dG) were
purchased from Boehringer-Mannheim. Restriction endo-
nuclease EcoRI and the Klenow fragment of DNA polymerase I
were purchased from New England Biolabs. EtBr and agarose
were purchased from Merck KgaA. TbCl3·6H2O was purchased
from Fluka Chemie and [α-32P]ATP was obtained from MP
Biomedicals. Melting points are uncorrected; NMR spectra
were run on a 500 MHz spectrometer; chemical shifts are
given in ppm (δ) and referenced relative to the internal solvent
signal; 195Pt NMR shifts are quoted relative to Ξ(195Pt) =
21.496784 MHz, K2PtCl4 was used as an external standard (δ =
−1612.81); mass spectra: direct inlet, EI, 70 eV; elemental ana-
lyses: Vario EL III elemental analyser; X-Ray diffractometer:
STOE-IPDS II; absorption and fluorescence measurements:
Tecan Infinite F200; absorption for DNA melting: Varian Cary
4000 UV-vis spectrophotometer; FAAS measurements: Varian
AA240Z Zeeman atomic absorption spectrometer equipped
with a GTA 120 graphite tube atomizer; flow cytometry:
Beckman Coulter Cytomics FC500. All biotested compounds
were >95% pure by elemental analysis. Benzimidazolium salts
were synthesised according to the literature procedures as
described in the ESI.† 12
cis-[Dichlorido(1,3-dimethylbenzimidazol-2-ylidene)(dimethyl-
sulfoxid)]platinum(II) (8a). As the solubility of 7a was not
sufficient to allow its separation from the by-products, 8a was
synthesised from 6a directly. Ag2O (106 mg, 456 µmol, 0.5 eq.)
was added to a solution of 6a (250 mg, 912 µmol) in CH2Cl2
(20 mL) and the mixture was stirred in the dark for 19 h. After
removal of the solvent, the residue was suspended in DMSO
(30 mL) and treated with K2PtCl4 (341 mg, 821 µmol, 0.9 eq.).
After 24 h at 60 °C in the dark, the reaction was worked up as
described above. Yield: 353 mg (720 µmol, 79%); colourless
crystals of melting point (mp) 262 °C; elemental analysis (%):
calc. for C11H16N2SOPtCl2 (490.31): C, 26.95; H, 3.29; N, 5.71.
1
Found: C, 27.24; H, 3.99; N, 6.26. H NMR (500 MHz, CDCl3):
δ 3.57 (6H, s, DMSO), 4.23 (6H, s, NCH3), 7.37 (2H, dd, J = 6.1,
3.1 Hz, BI), 7.44 (2H, dd, J = 6.1, 3.1 Hz, BI); 13C NMR
(125 MHz, CDCl3): δ 34.7 (NCH3), 46.3 (DMSO), 110.9 (BI),
124.3 (BI), 134.2 (BI-Cq), 154.8 (NHC); 195Pt NMR (108 MHz,
CDCl3): δ −3565.9; m/z (EI): 491 [M+], 413, 377, 340, 78.
cis-[Dichlorido-(1,3-diethylbenzimidazol-2-ylidene)(dimethyl-
sulfoxid)]platinum(II) (8b). 7b (449 mg, 1.41 mmol), K2PtCl4
(587 mg, 1.41 mmol, 1.0 eq.), DMSO (40 mL); yield: 639 mg
(1.23 mmol, 87%); colourless crystals of mp 270 °C; elemental
Syntheses and characterisation
Synthesis of NHC silver complexes 7. Solutions of the benzi- analysis (%): calc. for C13H20N2SOPtCl2 (518.37): C, 30.12; H,
midazolium chlorides in CH2Cl2 were treated with Ag2O 3.89; N, 5.40. Found: C, 30.75; H, 4.203; N, 5.96. 1H NMR
(0.5 eq.) and stirred for 24 h to 3 days in the dark. After fil- (500 MHz, CDCl3): δ 1.63 (6H, t, J = 7.3 Hz, CH3), 3.57 (6H, s,
tration, the product complexes 7 were precipitated by addition DMSO), 4.83 (4H, q, J = 7.3 Hz, CH2), 7.35 (2H, dd, J = 6.1, 3.1
of Et2O or hexane.
Hz, BI), 7.46 (2H, dd, J = 6.1, 3.1 Hz, BI); 13C NMR (125 MHz,
Chlorido(1,3-diethylbenzimidazol-2-ylidene) silver(I) (7b). 6b CDCl3): δ 14.5 (CH3), 43.8 (CH2), 46.4 (DMSO), 111.4 (BI), 124.0
(400 mg, 1.90 mmol), Ag2O (950 mg, 0.95 mmol, 0.5 eq.), (BI), 133.4 (BI-Cq), 153.5 (NCN); 195Pt NMR (108 MHz, CDCl3):
CH2Cl2 (40 mL), 24 h; yield: 526 mg (1.66 mmol, 87%); colour- δ −3561.4; m/z (EI): 518 [M+], 440, 404, 173, 78. Crystal data:
1
ˉ
less solid; H NMR (500 MHz, CDCl3): δ 1.49 (6H, t, J = 7.2 Hz, C13H20N2Cl2OPtS, M = 518.36, triclinic, space group P1,
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17375

View Article Online
Paper
Dalton Transactions
a = 8.675(5) Å, b = 9.264(5) Å, c = 10.601(5) Å, α = 91.874(5)°, J = 7.27 Hz, NCN); 31P NMR (CDCl3): δ 8.83 (J = 3893 Hz); 195Pt
β = 103.182(5)°, γ = 94.722(5)°, V = 825.5(8) Å3, Z = 2, NMR (CDCl3): δ −4017 (d, J = 3895 Hz); m/z (EI): 675 [M+], 640,
λ = 0.71069 Å, μ = 8.945 mm−1, T = 133 K; 7449 reflections 603, 263, 184, 148, 109.
measured, 3177 unique; R[F2 > 2σ(F2)] = 0.0329 and Rw = 0.0851,
GOF = 1.032. CCDC 1857512.†
cis-[Dichlorido(1,3-diethylbenzimidazol-2-ylidene)(triphenylphos-
phane)]platinum(II) (9b). Triphenylphosphane (131 mg,
cis-[Dichlorido-(1,3-di-n-butylbenzimidazol-2-ylidene)(dimethyl- 500 µmol, 5 eq.) was added to a solution of complex 8b
sulfoxid)]platinum(II) (8c). 7c (143 mg, 382 µmol), K2PtCl4 (50 mg, 97 µmol) in CH2Cl2 (5 mL) and the resulting mixture
(158 mg, 382 µmol, 1.0 eq.), DMSO (12 mL); yield: 203 mg was stirred at room temperature for 21 h. After precipitation
(353 µmol, 93%); colourless crystals of mp 227 °C; elemental from CH2Cl2/hexane the product was purified using column
analysis (%): calc. for C17H28NSOPtCl2 (574.47): C, 30.12; H, chromatography (silica gel, 1% MeOH in CH2Cl2). Yield: 29 mg
3.89; N, 5.40. Found: C, 30.75; H, 4.20; N, 5.96. 1H NMR (41 µmol, 43%); colourless solid of mp 323 °C. Elemental ana-
(500 MHz, CDCl3): δ 1.04 (6H, t, J = 7.4 Hz, CH3), 1.60–1.51 lysis (%): calc. for C29H29N2Cl2PPt (702.52): C, 49.58; H, 4.16;
1
(4H, m, NCCCH2), 2.04–1.94 (2H, m, NCCH2), 2.19–2.09 (2H, N, 3.99. Found: C, 49.19; H, 4.10; N, 3.70. H NMR (500 MHz,
m, NCCH2), 3.56 (6H, s, DMSO), 4.64 (2H, ddd, J = 14.0, 10.2, CDCl3): δ 1.46 (6H, t, J = 7.2 Hz, CH3), 3.84 (2H, dq, J = 14.5,
6.0 Hz, NCH2), 4.75 (2H, ddd, J = 14.1, 10.3, 5.6 Hz, NCH2), 7.2 Hz, NCH2), 4.85 (2H, dq, J = 14.5, 7.2 Hz, NCH2), 7.12 (2H,
7.33 (2H, dd, J = 6.1, 3.1 Hz, BI), 7.44 (2H, d, J = 9.2 Hz, BI); dd, J = 6.1, 3.1 Hz, BI), 7.24–7.16 (8H, m, BI, o-PPh3), 7.31 (3H,
13C NMR (125 MHz, CDCl3): δ 14.0 (CH3), 20.5 (CH3–CH2), 31.4 t, J = 7.2 Hz, p-PPh3), 7.64–7.53 (6H, m, m-PPh3); 13C NMR
(NCCH2), 46.3 (DMSO), 48.6 (NCH2), 111.4 (BI), 123.9 (BI), (125 MHz, CDCl3): δ 14.0 (CH3), 43.4 (NCH2), 110.3 (BI), 123.0
133.8 (BI-Cq), 153.7 (NCN); 195Pt NMR (CDCl3): δ −3554.2; m/z (BI), 128.2 (d, J = 10.9 Hz, o-PPh3), 129.0 (d, J = 64.5 Hz, ipso-
(EI): 574 [M+], 538, 501, 421, 229, 78.
PPh3), 131.0 (d, J = 2.72 Hz, p-PPh3), 133.3 (BI-Cq), 133.9 (d, J =
cis-[Dichlorido-(1,3-dioctylbenzimidazol-2-ylidene)(dimethyl- 10.9 Hz, m-PPh3), 160.6 (d, J = 7.26 Hz, NCN); 31P NMR
sulfoxid)]platinum(II) (8d). 7d (230 mg, 473 µmol), K2PtCl4 (CDCl3): δ 8.85 (J = 3861 Hz); 195Pt NMR (CDCl3): δ −4016 (d,
(196 mg, 473 µmol, 1.0 eq.), DMSO (15 mL); yield: 295 mg J = 3873 Hz); m/z (EI): 702 [M+], 666, 630, 455, 262.
(430 µmol, 91%); colourless solid of mp 132 °C; elemental ana-
cis-[Dichlorido(1,3-di-n-butylbenzimidazol-2-ylidene)(triphenyl-
lysis (%): calc. for C25H44N2Cl2SOPt (686.68): C, 43.73; H, 6.46; phosphane)]platinum(II) (9c). Triphenylphosphane (114 mg,
1
N, 4.08. Found: C, 44.36; H, 6.77; N, 4.36. H NMR (500 MHz, 434 µmol, 5 eq.) was added to a solution of complex 8c (50 mg,
CDCl3): δ 0.89 (6H, t, J = 6.87 Hz, CH3), 1.25–1.36 (12H, m, 86.9 µmol) in CH2Cl2 (5 mL) and the resulting mixture was
CH2), 1.36–1.44 (4H, m, NCCCH2), 1.48–1.56 (4H, m, stirred at room temperature for 30 min. The volatiles were
NCCCH2), 1.94–2.07 (2H, m, NCCH2), 2.09–2.20 (2H, m, removed in vacuo and the remaining crude product was puri-
NCCH2), 3.56 (6H, s, DMSO), 4.58–4.68 (2H, m, NCH2), fied by crystallisation from CH2Cl2/hexane. Yield: 48 mg
4.71–4.81 (2H, m, NCH2), 7.31–7.38 (2H, m, BI-H), 7.42–7.47 (63.3 µmol, 73%); colourless crystalline solid of mp 248 °C;
(2H, m, BI-H); 13C NMR (125 MHz, CDCl3): δ 14.1 (CH3), 22.6 elemental analysis (%): calc. for C33H37N2Cl2PPt (758.62): C,
(CH2), 27.1 (CH2), 29.2 (CH2), 29.3 (CH2), 29.3 (CH2), 31.6 52.25; H, 4.92; N, 3.69. Found: C, 52.75; H, 4.69; N, 3.58. 1H
(NCCH2), 46.2 (DMSO), 48.72 (NCH2), 111.3 (BI), 123.8 (BI), NMR (500 MHz, CDCl3): δ 0.94 (6H, t, J = 7.3 Hz, CH3),
133.6 (Cq), 153.5 (NHC); 195Pt NMR: δ −3556; m/z (EI): 686 1.35–1.48 (4H, m, NCCCH2), 1.63–1.73 (2H, m, NCCH2),
[M+], 533, 358, 230, 131, 78.
Synthesis of cis-[Cl2(PPh3)(1,3-dialkylbenzimidazol-2- (2H, m, NCH2), 7.11 (2H, dd, J = 6.0, 3.2 Hz, BI), 7.15–7.24 (8H,
ylidene]PtII complexes 9
m, BI and PPh3), 7.33 (3H, t, J = 7.3 Hz, PPh3), 7.48–7.66 (6H,
2.00–2.14 (2H, m, NCCH2), 3.78–3.87 (2H, m, NCH2), 4.54–4.64
cis-[Dichlorido(1,3-dimethylbenzimidazol-2-ylidene)(triphenyl- m, PPh3); 13C NMR (125 MHz, CDCl3): δ 13.7 (CH3), 20.4
phosphane)]platinum(II) (9a). A solution of triphenylphosphane (NCCCH2), 30.7 (NCCH2), 48.3 (NCH2), 110.4 (BI), 123.0 (BI),
(29 mg, 112 µmol, 1.1 eq.) in CH2Cl2 (15 mL) was added over 128.2 (d, J = 10.9 Hz, o-PPh3), 129.1 (d, J = 62.7 Hz, ipso-PPh3),
15 min to a solution of complex 8a (50 mg, 102 µmol) in 131.1 (d, J = 1.82 Hz, p-PPh3), 133.7 (BI-Cq), 134.0 (d, J = 10.9
CH2Cl2 (30 mL) and the resulting mixture was stirred at room Hz, m-PPh3), 160.7 (d, J = 7.26 Hz, NCN); 31P NMR (CDCl3):
temperature for a further 15 min. The volatiles were removed δ 9.19 (J = 3901 Hz); 195Pt NMR: δ −4018 (d, J = 3871 Hz); m/z
in vacuo and the remaining crude product was purified using (EI): 758 [M+], 722, 685, 262, 229, 183, 108. Crystal data:
ˉ
column chromatography (silica gel, 2% MeOH in CH2Cl2). The C33H37Cl2N2PPt, M = 758.63, triclinic, space group P1, a =
product was crystallised from CH2Cl2/Et2O. Yield: 22 mg 9.1067(18) Å, b = 12.792(3) Å, c = 13.803(3) Å, α = 88.02(3)°, β =
(32.6 µmol, 32%); colourless crystalline solid of mp 346 °C. 87.59(3)°, γ = 71.06(3)°, V = 1519.2(6) Å3, Z = 2, λ = 0.71073 Å,
Elemental analysis (%): calc. for C27H25N2Cl2PPt (674.46): C, μ = 4.872 mm−1, T = 133 K; 21 191 reflections measured, 5813
48.08; H, 3.74; N, 4.15. Found: C, 47.71; H, 4.03; N, 4.14. 1H unique; R[F2 > 2σ(F2)] = 0.0183 and Rw = 0.0447, GOF = 1.030.
NMR (500 MHz, CDCl3): δ 3.83 (6H, s, CH3), 7.07–7.15 (2H, m, CCDC 1857511.†
BI), 7.16–7.26 (8H, m, BI and PPh3), 7.29–7.37 (3H, m, PPh3),
cis-[Dichlorido(1,3-dioctylbenzimidazol-2-ylidene)(triphenylpho-
7.60 (6H, t, J = 9.5 Hz, PPh3); 13C NMR (125 MHz, CDCl3): sphane)]platinum(II) (9d). Triphenylphosphane (95 mg,
δ 34.1 (CH3), 109.9 (BI), 123.1 (BI), 128.29 (d, J = 11.8 Hz, 364 µmol, 5 eq.) was added to a solution of complex 8d
o-PPh3), 129.4 (d, J = 63.5 Hz, ipso-PPh3), 131.1 (d, J = 1.82 Hz, (50 mg, 73 µmol) in CH2Cl2 (20 mL) and the resulting mixture
p-PPh3), 134.0 (d, J = 10.9 Hz, m-PPh3), 134.1 (BI-Cq), 162.0 (d, was stirred at room temperature for 24 h. After crystallisation
17376 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
from CH2Cl2/hexane the crude product was purified using J = 7.2 Hz, CH2), 7.20–7.42 (24H, m, BI and PPh3), 7.43–7.54
column chromatography (silica gel, 1% MeOH in CH2Cl2). (10H, m, PPh3); 13C NMR (125 MHz, CDCl3): δ 13.0 (CH3), 44.2
Yield: 60 mg (68 µmol, 93%); colourless crystalline solid of mp (CH2), 111.5 (BI), 124.6 (BI), 127.1 (t, J = 25.5 Hz, ipso-PPh3),
168 °C. Elemental analysis (%): calc. for C41H53N2Cl2PPt 129.0 (t, J = 5.5 Hz, o-PPh3), 132.0 (p-PPh3), 133.5 (BI-Cq), 133.7
(870.84): C, 56.55; H, 6.13; N, 3.22. Found: C, 57.14; H, 6.03; N, (m-PPh3), 158.0 (t, J = 9.5 Hz, NCN); 31P NMR (CDCl3): δ 18.9
3.22. 1H NMR (500 MHz, CDCl3): δ 0.91 (6H, t, J = 6.9 Hz, CH3), (J = 2485 Hz); 195Pt NMR: δ −4383 (t, J = 2490 Hz); m/z (EI):
1.26–1.39 (20H, m, CH2), 1.63–1.72 (2H, m, NCCH2), 2.04–2.13 702, 667, 630, 456, 297, 262, 183, 108.
(2H, m, NCCH2), 3.78–3.88 (2H, m, NCH2), 4.52–4.78 (2H, m,
trans-[Chlorido(1,3-di-n-butylbenzimidazol-2-ylidene)bis(triphe-
NCH2), 7.10 (2H, dd, J = 6.9, 3.05 Hz, BI), 7.16–7.25 (8H, m, BI nylphosphane)]platinum(II) chloride (10c). Triphenylphosphane
and PPh3), 7.33 (3H, t, J = 6.7 Hz, PPh3), 7.51–7.64 (6H, m, (456 mg, 1.74 mmol, 20 eq.) was added to a solution of
PPh3); 13C NMR (125 MHz, CDCl3): δ 14.1 (CH3), 22.6 (CH2), complex 8b (50 mg, 87 µmol) in CH2Cl2 (10 mL) and the result-
27.3 (CH2), 28.8 (CH2), 29.1 (NCCH2), 29.3 (CH2), 31.6 (CH2), ing mixture was stirred at 80 °C for 4 days. The volatiles were
48.6 (NCH2), 110.4 (BI), 122.9 (BI), 128.1 (d, J = 11.8 Hz, removed in vacuo and the remaining crude product was puri-
o-PPh3), 129.2 (d, J = 62.7 Hz, ipso-PPh3), 131.0 (d, J = 2.7 Hz, fied by crystallisation from CH2Cl2/Et2O and subsequent
p-PPh3), 134.0 (d, J = 10.9 Hz, m-PPh3), 160.7 (d, J = 8.17 Hz, column chromatography (silica gel, CH2Cl2 with 2–5% MeOH).
NCN); 31P NMR (CDCl3): δ 9.17 (J = 3874 Hz); 195Pt NMR Yield: 29 mg (28 µmol, 33%); colourless solid of mp 114 °C;
(CDCl3): δ −4018 (d, J = 3879 Hz); m/z (EI): 870 [M+], 834, 797, elemental analysis (%): calc. for C51H52N2Cl2P2Pt (1020.91): C,
564, 533, 341, 262, 183, 108.
Synthesis of trans-[Cl(1,3-dialkylbenzimidazol-2-ylidene) NMR (500 MHz, CDCl3): δ 0.70 (6H, t, J = 7.0 Hz, CH3),
(PPh3)2] Pt+ Cl− complexes 10
1.12–1.23 (8H, m, CH2), 3.62–3.70 (4H, m, NCH2), 7.15 (2H,
60.00; H, 5.13; N, 2.74. Found: C, 59.26; H, 5.11; N, 3.06. 1H
trans-[Chlorido(1,3-dimethylbenzimidazol-2-ylidene)bis(triphenyl- dd, J = 6.5, 3.1 Hz, BI), 7.18–7.42 (22H, m, BI and PPh3),
phosphane)]platinum(II) chloride (10a). Triphenylphosphane 7.42–7.57 (10H, m, PPh3); 13C NMR (125 MHz, CDCl3): δ 13.3
(80 mg, 306 µmol, 3 eq.) was added to a solution of complex (CH3), 20.4 (CH2CH3), 30.2 (NCCH2), 49.1 (CH2), 111.2 (BI),
8a (50 mg, 102 µmol) in CH2Cl2 (5 mL) and the resulting 124.8 (BI), 127.2 (t, J = 29.1 Hz, ipso-PPh3), 129.0 (t, J = 5.5 Hz,
mixture was stirred at room temperature for 30 min. The vola- o-PPh3), 132.1 (p–PPh3), 133.6 (BI-Cq), 133.9 (m-PPh3), 158.4 (t,
tiles were removed in vacuo and the remaining crude product J = 10.0 Hz, NCN); 31P NMR (CDCl3): δ 18.9 (J = 2486 Hz); 195Pt
was purified by crystallisation from CH2Cl2/hexane. Yield: NMR: δ −4379 (t, J = 2486 Hz); m/z (EI): 758, 723, 685, 456, 262,
80 mg (90 µmol, 83%); colourless crystalline solid of mp 229, 183, 108.
310 °C; elemental analysis (%): calc. for C48H47N2P2PtCl2
X-ray data collection and structural determination
(979.84): C, 57.70; H, 4.30; N, 2.99. Found: C, 57.98; H, 5.04; N,
1
3.43. H NMR (500 MHz, CDCl3): δ 3.54 (6H, s, CH3), 7.20 (2H, Diffractometer used: STOE-STADIVARI; data collection and cell
dd, J = 6.1, 3.1 Hz, BI), 7.29 (12H, t, J = 7.5 Hz, o–PPh3), 7.32 refinement by: STOE-X-AREA. The single crystal samples were
(2H, dd, J = 6.1, 3.1 Hz, BI), 7.37 (6H, t, J = 7.4 Hz, p–PPh3), measured with Mo-Kα and the reflections collected at 133 K.
7.51 (12H, dd, J = 12.4, 6.0 Hz, m-PPh3); 13C NMR (125 MHz, The structures were solved by direct methods using SIR97 and
CDCl3): δ 35.1 (CH3), 111.1 (BI), 124.3 (BI), 127.5 (t, J = 29.5 Hz, refined by full matrix least-squares on F2 for all data using
ipso-PPh3), 128.9 (t, J = 5.5 Hz, o-PPh3), 131.8 (p-PPh3), 133.7 (t, SHELXL2016/6. All hydrogen atoms were added at calculated
J = 5.9 Hz, m-PPh3), 133.9 (BI-Cq), 159.9 (t, J = 10.0 Hz, NCN); positions and refined using a riding model. Anisotropic
31P NMR (CDCl3): δ 17.7 (J = 2478 Hz); 195Pt NMR: δ −4378 (t, thermal displacement parameters were used for all non-hydro-
J = 2490 Hz); m/z (EI): 908 [M+], 673, 653, 602, 262, 183. Crystal gen atoms. Further details of the data collection and reliability
data: C45H40ClN2P2Pt·Cl, M = 936.72, monoclinic, space group factors are listed together with the analytical data in the ESI
P21/c, a = 12.578 (5) Å, b = 10.790 (5) Å, c = 32.126 (5) Å, α = (Table S1†). Supplementary crystallographic data were de-
90°, β = 94.531(5)°, γ = 90°, V = 4346 (3) Å3, Z = 4, λ = 0.71069 Å, posited with The Cambridge Crystallographic Data Centre
μ = 3.46 mm−1, T = 133 K; 32 457 reflections measured, 6540 CCDC under no. 1857512 (8b), 1857511 (9c), 1857513 (10a).†
unique; R[F2 > 2σ(F2)] = 0.0276 and Rw = 0.0601, GOF = 0.766.
Biological evaluation
CCDC 1857513.†
trans-[Chlorido(1,3-diethylbenzimidazol-2-ylidene)bis(triphenyl-
phosphane)]platinum(II) chloride (10b). Triphenylphosphane 518A2
Cell culture conditions and stock solutions. Cells of
melanoma, HT-29, DLD-1, and
HCT116wt
(252 mg, 960 µmol, 10 eq.) was added to a solution of complex HCT116p53−/− colon carcinoma, MCF-7Topo mamma carci-
8b (50 mg, 96 µmol) in CH2Cl2 (10 mL) and the resulting noma, and Kb-V1Vbl cervix carcinoma were kept in DMEM. All
mixture was stirred at room temperature for 5 d. The volatiles cell lines were cultured at 37 °C, 95% humidity and 5% CO2.
were removed in vacuo and the remaining crude product was To maintain the mdr characteristics of the MCF-7Topo and
purified by crystallisation from CH2Cl2/Et2O. Yield: 22 mg Kb-V1Vbl cell lines, they were regularly treated with either topo-
(22.8 µmol, 24%); colourless solid of mp 297 °C; elemental tecan or vinblastine. Dilutions of platinum(II) complexes 8, 9
analysis (%): calc. for C50H51N2Cl2P2Pt (1007.89): C, 58.51; H, and 10a–c were made from 10 mM stock solutions (DMSO) in
4.60; N, 2.90. Found: C, 58.59; H, 4.66; N, 2.89. 1H NMR sterile ddH2O, dilutions of CDDP from a 3.3 mM stock solu-
(500 MHz, CDCl3): δ 0.83 (6H, t, J = 7.3 Hz, CH3), 3.86 (4H, q, tion in 150 mM saline.25 Plasmid pSP73 (2464 bp) was isolated
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17377

View Article Online
Paper
Dalton Transactions
according to standard procedures and banded twice in CsCl highest concentration. To determine the background fluo-
equilibrium density gradients. rescence, samples were prepared analogously with equal
Growth inhibition assay (MTT-assay). For the investigation amounts of TE buffer instead of the DNA stock solution.
of the cytotoxicity an MTT (3-(4,5-dimethylthiazole-2-yl)-2,5- 100 µL of ethidium bromide in TE-buffer (10 µg mL−1) was
diphenyltetrazoliumbromide) based proliferation assay was added to each well and the plate was incubated in the dark for
performed. Cells were seeded into 96-well plates (0.05 × 106 5 min at room temperature. The ethidium bromide fluo-
cells per mL for 518A2, HT-29, HCT116, MCF-7Topo and rescence was measured at λex = 535 nm and λem = 590 nm. The
Kb-V1Vbl, and 0.1 × 106 cells per mL for DLD-1) and incubated background signal was subtracted from the respective sample
for 24 h under cell culture conditions to ensure their adhesion. signals and the fluorescence intensities relative to the control
On the next day the cells were treated with a series of appropri- (set to 100%) were calculated. Each measurement was run in
ate dilutions (final concentrations ranging from 5 nM, resp. triplicate. Reduction of the fluorescence indicates an inter-
5 pM, to 100 µM) of 8, 9, 10a–c and CDDP. As a negative action of 8, 9, 10a–c or CDDP with the DNA and thus a hin-
control (100% vital cells) the assay was performed analogously drance of ethidium bromide intercalation.
with the respective vehicle. The cells were incubated for a
Electrophoretic mobility shift assay (EMSA). Circular
further 72 h under cell culture conditions, centrifuged (300g, pBR322 plasmid DNA (1.5 µg) in TE buffer was incubated for
4 °C) and the media were exchanged for 50 µL of a 0.05% MTT 24 h with 0, 5, 10, 25 and 50 µM of complexes 8, 9 or 10a–c at
solution in PBS. The cells were incubated for another 2 h at 37 °C (final sample volume: 20 µL). On the next day DNA
37 °C before the plates were again centrifuged as before and sample buffer (5×, Tris/HCl, pH 8.0; 25% glycerol) was added
the MTT solution was discarded. The water insoluble formazan to each sample and an agarose gel electrophoresis (1% in TBE
was dissolved by addition of 25 µL of an SDS/DMSO solution buffer; 90 mM Tris/HCl, pH 8.3; 90 mM boric acid; 2.5 mM
(10%; 0.6% AA) to each well. After 1 h of incubation at 37 °C EDTA) was run at 66 V for 4 h. The DNA was stained with ethi-
the absorption was measured at 570 nm, and the background dium bromide and the plasmid bands were visualised via a UV
absorption at 630 nm. The formazan absorption correlates transilluminator.
directly with the number of viable cells. The IC50 values were
determined with GraphPad Prism and means SD were calcu- incubated with 8c, 9c or 10a–c (100 µM) in 10 mM NaClO4 in
lated from six independent measurements. the dark at 37 °C. After 2, 4 or 24 h, aliquots were withdrawn
DNA platination analysis. Calf thymus DNA (1 mM) was
Inhibition of P-gp1. The determination of the cytotoxicity of and exhaustively dialysed against water or, alternatively,
8b–d, 9b–d and 10a–c in Kb-V1Vbl cervix carcinoma cells in the against 0.1 M NaCl and subsequently against water. The
presence of verapamil, a competitive inhibitor of P-gp1, was amount of platinum associated with the DNA was assessed
conducted analogously to the normal MTT-assay, but with using FAAS. The percentage of Pt bound to DNA was calculated
DMEM containing 24 µM of verapamil (10 mM stock solution as the ratio of Pt associated with DNA and the total amount of
in 70% EtOH).
platinum present in the sample.
Cellular uptake. The uptake of complexes 8a–c, 9a–d, 10a–c
DNA melting point analysis. DNA was modified with Pt com-
and CDDP by HCT116wt cells was measured via ICP-MS. The plexes for 24 h in 10 mM NaClO4 at 37 °C. The DNA was di-
cells were seeded on 100 mm tissue culture dishes (3 × 106 alysed against 0.1 M NaCl to remove the unbound platinum
cells per dish in 7 mL of growth medium) and incubated at and subsequently against the medium required for further
37 °C. Confluent cells were treated with the compounds at analysis. An aliquot of the sample was used to determine the
final equimolar concentrations (8 µM) for 5 h. The attached level of DNA platination (rb) using FAAS. The DNA melting
cells were harvested by trypsinisation and the viability of the curves of unplatinated or platinated DNA were recorded by
cells was tested by a trypan blue exclusion assay. Viabilities of measuring the absorbance at 260 nm in a medium containing
the cells ranged from 76 to 97%, both viable and dead frac- 10 mM or 0.1 M Na+ with 1 mM Tris-HCl/0.1 mM EDTA, pH 7.4,
tions were used for the experiment. The cell pellets were sub- using a Varian Cary 4000 UV-vis spectrophotometer equipped
sequently washed twice with cold PBS and cells were counted with a thermoelectrically controlled cell holder and quartz
by using a TC10 Automated Cell Counter (Bio-Rad). The cells cells with a pathlength of 1 cm. The melting temperature (Tm)
were digested by using a microwave acid (HCl, 11 M) digestion was determined as the temperature corresponding to
a
system (MARS5, CEM) to give a fully homogenised solution. maximum on the first derivation profile of the melting curves.
The final platinum content in the samples was determined
using ICP-MS.
Tb3+ fluorescence. Ct DNA was modified with Pt complexes
in 10 mM NaClO4 at 37 °C. After 24 h of incubation, DNA was
dialysed against 0.1 M NaCl and then against 10 mM NaClO4.
The level of DNA platination was verified using FAAS. The
DNA interaction
Ethidium bromide saturation assay. Inside the wells of a terbium fluorescence was measured using a previously described
black 96-well plate 1 µg salmon sperm DNA in 100 µL of method.18 Briefly, the samples of platinated or unplatinated
freshly sterile-filtered TE buffer (10 mM Tris/HCl, 1 mM EDTA DNA (2.5 × 10−5 M) were incubated with 5 × 10−5 M TbCl3 for
pH 8.5) was incubated with 5, 10, 25 and 50 µM of 8, 9, 10a–c 60 min at 25 °C. Subsequently, the fluorescence intensity was
or CDDP for 2 h at 37 °C (final sample volume 111.1 µL). measured in a 1 cm quartz cell using a Varian Cary Eclipse
Control samples were prepared analogously with DMSO to the spectrofluorophotometer. Excitation and emission wavelengths
17378 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
were set to 290 and 546 nm, respectively, the slit widths were respective tubes, the cells were trypsinised for ca. 2 min at
5 nm, and the integration time was set to 5 s. 37 °C. The detached cells were transferred into the tubes and
Unwinding of negatively supercoiled DNA. The unwinding centrifuged (300g, 4 °C, 5 min). The supernatant was discarded
of closed circular supercoiled pSP73 plasmid DNA was assayed and the cells were fixed in 1 mL of ice-cold 70% EtOH for 1 h
using an agarose gel mobility shift assay.19 Plasmid DNA at 4 °C.
(pSP73) was incubated with the platinum complex for 24 h,
For propidium iodide staining, the cells were centrifuged
precipitated by ethanol to remove unbound Pt, and redissolved (400g, room temperature, 5 min), the supernatant was dis-
in TAE buffer (0.04 M Tris-acetate, 1 mM EDTA, pH 7.0). An carded, and the pellet was layered with 1 mL PBS. After 5 min
aliquot of the precipitate was subjected to electrophoresis on the cells were again centrifuged and the PBS discarded. The
1% agarose gels running at 25 °C in the dark with TAE buffer. cells were resuspended in 200 µL of a PI staining solution
The gels were stained with EtBr, followed by photography with (0.1% sodium citrate, 50 µg mL−1 propidium iodide, 50 µg mL−1
a transilluminator. Another aliquot was used to determine the RNase A) for 30 min at 37 °C. The distribution of the cells in
rb values using FAAS. The unwinding angle θ, induced per the different phases of the cell cycle was measured via flow
platinum–DNA adduct, was calculated based upon the rb value cytometry.
at which the transformation of the supercoiled to a relaxed
form of the plasmid was complete.19
Interstrand cross-link formation assay. The complexes 8c, Conflicts of interest
9c, 10a–c and CDDP were incubated at different concentrations
(CDDP rb = 0, 0.001, 0.0015, 0.002; 8c, 9c, 10c rb = 0, 0.01,
There are no conflicts to declare.
0.015, 0.02) for 24 h with 500 ng of pSP73KB DNA after it had
been linearized by EcoRI and 3′-end-labeled by means of a
Klenow fragment of DNA polymerase I and [α-32P]dATP. After
Acknowledgements
incubation, the reaction was stopped by addition of NaCl to a
RS thanks the Deutsche Forschungsgemeinschaft for a grant
final concentration of 0.1 M and samples were precipitated.
(Scho 402/12), JK, HK and VB thank the Czech Science
The sediments were dissolved in 10 mM NaClO4, and the
Foundation and Ministry of Education of the Czech Republic
number of interstrand crosslinks was analysed using electro-
for grant (18-09502S and LTC17003), and MR thanks the
phoresis under denaturing conditions on an alkaline agarose
Bayreuth University Graduate School for financial support. We
also thank the COST Action CM1105 ‘Functional metal com-
plexes that bind to biomolecules’ for financial support.
gel (1%). After the electrophoresis had been completed, the
gels were dried and visualised using a FUJIFILM BAS 2500 bio-
imaging analyser. The intensities of the bands corresponding
to single strands of DNA and interstrand cross-linked duplex
were quantified using AIDA image analysis software. The fre-
quency of ICLs, was calculated as % ICL/Pt = XL/4910 × rb (the
References
pSP73KB plasmid contained 4910 nucleotide residues), in
which % ICL/Pt is the number of ICLs per adduct and XL is
the number of ICLs per molecule of the linearized DNA duplex
and was calculated assuming a Poisson distribution of the
ICLs as XL = −ln A. In which A is the fraction of molecules
running as a band corresponding to the non-crosslinked DNA.
The values of rb were determined using FAAS by measuring the
amount of free, unbound platinum in the supernatants after
precipitation of the samples; the amount of platinum bound
to DNA, rb, was calculated by subtracting the amount of plati-
num remaining in the solution from the total amount of plati-
num present in the reaction.
Cell cycle analysis. The cell cycle analysis of the HCT116
cells was performed via flow cytometry of the treated HCT116wt
and HCT116−/− cells. They were seeded at 0.1 × 106 cells per mL
into the wells of a 6-well plate and incubated for 24 h
under cell culture conditions. Appropriate dilutions of 8c, 9c,
10c and CDDP in ddH2O (30.3 µL) were added to each well
(final concentrations: 8c 5 µM, 9c and 10c 500 nM, CDDP
10 µM). The cells were incubated for another 24 h under cell
culture conditions. On the next day the medium from each
well was transferred into a centrifugation tube and the cells
were washed with 1 mL PBS. After transferring the PBS to the
1 (a) A.-M. Florea and D. Büsselberg, Cancers, 2011, 3, 1351–
1371, DOI: 10.3390/cancers3011351; (b) S. Dasari and
P. B. Tchounwou, Eur. J. Pharmacol., 2014, 740, 364–378,
DOI: 10.1016/j.ejphar.2014.07.025.
2 (a) K. M. Deo, D. L. Ang, B. McGhie, A. Rajamanickam,
A. Dhiman, A. Khoury, J. Holland, A. Bjelosevic, B. Pages,
C. Gordon and J. R. Aldrich-Wright, Coord. Chem. Rev.,
2018, 375, 148–163, DOI: 10.1016/j.ccr.2017.11.014;
(b) T. C. Johnstone, K. Suntharalingam and S. J. Lippard,
Chem. Rev., 2016, 116, 3436–3486, DOI: 10.1021/acs.
chemrev.5b00597.
3 A. J. Arduengo, R. L. Harlow and M. Kline, J. Am. Chem.
Soc., 1991, 113, 361–363, DOI: 10.1021/ja00001a054.
4 X. Bantreil, J. Broggi and S. P. Nolan, Annu. Rep. Prog.
Chem., Sect. B, 2009, 105, 232–263, DOI: 10.1039/b822056p.
5 (a) L. Oehninger, R. Rubbiani and I. Ott, Dalton Trans.,
2013, 42, 3269–3284, DOI: 10.1039/c2dt32617e; (b) W. Liu
and R. Gust, Coord. Chem. Rev., 2016, 329, 191–213, DOI:
10.1016/j.ccr.2016.09.004.
6 (a) S. Patil, A. Deally, B. Gleeson, H. Müller-Bunz,
F. Paradisi and M. Tacke, Metallomics, 2011, 3, 74–88, DOI:
10.1039/c0mt00034e; (b) R. A. Haque, S. Y. Choo,
S. Budagumpi, M. A. Iqbal and A. Al-Ashraf Abdullah,
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17379

View Article Online
Paper
Dalton Transactions
Eur. J. Med. Chem., 2015, 90, 82–92, DOI: 10.1016/j.
ejmech.2014.11.005.
7 (a) R. Rubbiani, S. Can, I. Kitanovic, H. Alborzinia,
M. Stefanopoulou, M. Kokoschka, S. Monchgesang,
C. Cararelli, et al., J. Biol. Chem., 1986, 261, 7762–7770;
(c) M. M. Cromwell, A. R. Safa, R. L. Felsted,
M. M. Gottesman and I. Pastan, Proc. Natl. Acad. Sci. U. S. A.,
1986, 83, 3847–8350, DOI: 10.1073/pnas.83.11.3847.
W. S. Sheldrick, S. Wölfl and I. Ott, J. Med. Chem., 2011, 54, 14 (a) P. Bragado, A. Armesilla, A. Silva and A. Porras, Apoptosis,
8646–8657, DOI: 10.1021/jm201220n; (b) J. K. Muenzner,
B. Biersack, H. Kalie, I. C. Andronache, L. Kaps,
D. Schuppan, F. Sasse and R. Schobert, ChemMedChem,
2014, 9, 1195–1204, DOI: 10.1002/cmdc.201400049;
(c) J. K. Muenzner, B. Biersack, A. Albrecht, T. Rehm,
U. Lacher, W. Milius, A. Casini, J. Zhang, I. Ott, V. Brabec,
O. Stuchlikova, I. C. Andronache, L. Kaps, D. Schuppan
and R. Schobert, Chem. – Eur. J., 2016, 22, 18953–18962,
DOI: 10.1002/chem.201604246.
2007, 12, 1733–1742, DOI: 10.1007/s10495-007-0082-8;
(b) Q. Kai, L. Ting, W. Jichao, M. Fandi, W. Zhixin,
H. Zichao, W. Yong, S. Sidong, L. Sinan, C. Hulin, D. Yafeng
and L. Chang, J. South. Med. Univ., 2013, 33, 1253–1259,
DOI: 10.3969/j.ssn.1673-4254.2013.09.01; (c) N. S. Pellegata,
R. J. Antoniono, J. L. Redpath and E. J. Standbridge, Proc.
Natl. Acad. Sci. U. S. A., 1996, 93, 15209–15214, DOI:
10.1073/pnas.93.26.15209.
15 G. He, Z. H. Siddik, Z. Huang, R. Wang, J. Koomen,
R. Kobayashi, A. R. Khokhar and J. Kuang, Oncogene, 2005,
24, 2929–2943, DOI: 10.1038/sj.onc.1208474.
8 (a) W. Streciwilk, F. Hackenberg, H. Müller-Bunz and
M. Tacke, Polyhedron, 2014, 80, 3–9, DOI: 10.1016/j.
poly.2013.11.039; (b) M.-L. Teyssot, A.-S. Jarrousse, 16 (a) S. Tabassum, G. Chandra Sharma, F. Arjmad and
A. Chevry, A. De Haze, C. Beaudoin, M. Manin, S. P. Nolan,
S. Díez-González, L. Morel and A. Gautier, Chem. – Eur. J.,
2009, 15, 314–318, DOI: 10.1002/chem.200801992.
A. Azam, Nanotechnology, 2010, 21, 195102–114094, DOI:
10.1088/0957-4484/21/19/195102; (b) V. Vijayanathan,
T. Thomas and T. J. Thomas, Biochemistry, 2002, 41, 14085–
14094, DOI: 10.1021/bi0203987.
9 (a) G. Lv, L. Guo, L. Qiu, H. Yang, T. Wang, H. Liu and
J. Lin, Dalton Trans., 2015, 44, 7324–7331, DOI: 10.1039/ 17 (a) N. P. Johnson, Biochem. Biophys. Res. Commun., 1982,
c5dt00169b; (b) L. Oehninger, M. Stefanopoulou,
H. Alborzinia, J. Schur, S. Ludewig, K. Namikawa,
A. Muñoz-Castro, R. W. Köster, K. Baumann, S. Wölfl,
W. S. Sheldrick and I. Ott, Dalton Trans., 2013, 42, 1657–
1666, DOI: 10.1039/c2dt32319b.
104, 1394–1400, DOI: 10.1016/0006-291X(82)91404-8;
(b) A. M. J. Fichtinger-Schepman, J. L. Van der Veer,
J. H. J. Den Hartog, P. H. M. Lohman and J. Reedijk,
Biochemistry, 1985, 24, 707–713, DOI: 10.1021/bi00324a025;
(c) A. Eastman, Pharmacol. Ther., 1987, 34, 155–166, DOI:
10.1016/0163-7258(87)90009-X.
10 (a) M. Skander, P. Retailleau, B. Bourrié, L. Schio,
P. Mailliet and A. Marinetti, J. Med. Chem., 2010, 53, 2146– 18 Z. Balcarova and V. Brabec, Biophys. Chem., 1989, 33, 55–
2154, DOI: 10.1021/jm901693m; (b) E. Chardon, G. Dahm, 61, DOI: 10.1016/0301-4622(89)80007-9.
G. Guichard and S. Bellemin-Laponnaz, Organometallics, 19 M. V. Keck and S. J. Lippard, J. Am. Chem. Soc., 1992, 114,
2012, 31, 7618–7621, DOI: 10.1021/om300806g; (c) R. Wai- 3386–3390, DOI: 10.1021/ja00035a033.
Yin Sun, A. Lok-Fung Chow, X.-H. Li, J. J. Yan, S. Sin-Yin 20 (a) G. L. Cohen, W. R. Bauer, J. K. Barton and S. J. Lippard,
Chui and C.-M. Che, Chem. Sci., 2011, 2, 728–736, DOI:
10.1039/c0sc00593b; (d) J. K. Muenzner, T. Rehm,
B. Biersack, A. Casini, I. de Graaf, P. Worawutputtapong,
Science,
1979,
203,
1014–1016,
DOI:
10.1126/
science.370979; (b) W. M. Scovell and F. Collart, Nucleic
Acids Res., 1985, 13, 2881–2895.
A. Noor, R. Kempe, V. Brabec, J. Kasparkova and 21 R. Zaludova, V. Kleinwächter and V. Brabec, Biophys. Chem.,
R. Schobert, J. Med. Chem., 2015, 58, 6283–6292, DOI: 1996, 60, 135–142, DOI: 10.1016/0301-4622(96)00010-5.
10.1021/acs.jmedchem.5b00896; (e) M. Bouché, G. Dahm, 22 (a) K. S. Lovejoy, M. Serova, I. Bieche, S. Emami,
A. Maisse-François, T. Achard and S. Bellemin-Laponnaz,
Eur. J. Inorg. Chem., 2016, 17, 2828–2836, DOI: 10.1002/
ejic.201600296.
M. D’Incalci, M. Broggini, E. Erba, C. Gespach,
E. Cvitkovic, S. Faivre, E. Raymond and S. J. Lippard, Mol.
Cancer Ther., 2011, 10, 1709–1719, DOI: 10.1158/1535-7163.
MCT-11-0250; (b) G. Y. Park, J. J. Wilson, Y. Song and
S. J. Lippard, Proc. Natl. Acad. Sci. U. S. A., 2012, 109,
11987–11992, DOI: 10.1073/pnas.1207670109; (c) B. Wang,
Z. Wang, F. Ai, W. K. Tang and G. Zhu, J. Inorg. Biochem.,
2015, 142, 118–125, DOI: 10.1016/j.jinorgbio.2014.10.003;
(d) G. Y. Zhu, M. Myint, W. H. Ang, L. Song and
S. J. Lippard, Cancer Res., 2012, 72, 790–800, DOI: 10.1158/
0008-5472.CAN-11-3151.
11 T. Rehm, M. Rothemund, J. K. Muenzner, A. Noor,
R. Kempe and R. Schobert, Dalton Trans., 2016, 45, 15390–
15398, DOI: 10.1039/c6dt02350a.
12 (a) R. Rubbiani, I. Kitanovic, H. Alborzinia, S. Can,
A. Kitanovic, L. A. Onambele, M. Stefanopoulou,
Y. Geldmacher, W. S. Sheldrick, G. Wolber, A. Prokop,
S. Wölfl and I. Ott, J. Med. Chem., 2010, 53, 8608–8618,
DOI: 10.1021/jm100801e; (b) H. Valdés, M. Poyatos,
G. Ujaque and E. Peris, Chem. – Eur. J., 2015, 21, 1578– 23 R. C. Todd and S. J. Lippard, Metallomics, 2009, 1, 280–291,
1588, DOI: 10.1002/chem.201404618; (c) H. Lu and DOI: 10.1039/b907567d.
R. L. Brutchey, Chem. Mater., 2017, 29, 1396–1403, DOI: 24 (a) S. Teletchéa, S. Komeda, J.-M. Teuben, M. A. Elizondo-
10.1021/acs.chemmater.6b05293.
13 (a) T. Nakanishi and D. D. Ross, Chin. J. Cancer, 2012, 31,
73–99, DOI: 10.5732/cjc.011.10320; (b) D. Shen,
Riojas, J. Reedijk and J. Kozelka, Chem. – Eur. J., 2006, 12,
3741–3753; (b) J. Mlcouskova, J. Malina, V. Novohradsky,
J. Kasparkova, S. Komeda and V. Brabec, Biochim. Biophys.
17380 | Dalton Trans., 2018, 47, 17367–17381
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
Acta,
2012,
1820,
1502–1511,
DOI:
10.1016/j. 25 M. D. Hall, K. A. Telma, K.-E. Chang, T. D. Lee,
bbagen.2012.05.014; (c) J. Mlcouskova, J. Kasparkova,
T. Suchankova, S. Komeda and V. Brabec, J. Inorg. Biochem.,
2012, 114, 15–23, DOI: 10.1016/j.jinorgbio.2012.04.015.
J. P. Madigan, J. R. Lloyd, I. S. Goldlust, J. D. Hoeschele
and M. M. Gottesman, Cancer Res., 2014, 74, 3913–3922,
DOI: 10.1158/0008-5472.CAN-14-0247.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans., 2018, 47, 17367–17381 | 17381