Rhodium(I) N-Heterocyclic Carbene Bioorganometallics as in Vitro Antiproliferative Agents with Distinct Effects on Cellular Signaling

Article abstract of DOI:10.1021/acs.jmedchem.5b01159

Organometallics with N-heterocyclic carbene (NHC) ligands have triggered major interest in inorganic medicinal chemistry. Complexes of the type Rh(I)(NHC)(COD)X (where X is Cl or I, COD is cyclooctadiene, and NHC is a dimethylbenzimidazolylidene) represent a promising type of new metallodrugs that have been explored by advanced biomedical methods only recently. In this work, we have synthesized and characterized several complexes of this type. As observed by mass spectrometry, these complexes remained stable over at least 3 h in aqueous solution, after which hydrolysis of the halido ligands occurred and release of the NHC ligand was evident. Effects against mitochondria and general cell tumor metabolism were noted at higher concentrations, whereas phosphorylation of HSP27, p38, ERK1/2, FAK, and p70S6K was induced substantially already at lower exposure levels. Regarding the antiproliferative activity in tumor cells, a clear preference for iodido over chlorido secondary ligands was noted, as well as effects of the substituents of the NHC ligand.

Full text of DOI:10.1021/acs.jmedchem.5b01159

Article
pubs.acs.org/jmc
Rhodium(I) N‑Heterocyclic Carbene Bioorganometallics as in Vitro
Antiproliferative Agents with Distinct Effects on Cellular Signaling
†
†
‡
§
‡
Luciano Oehninger, Sarah Spreckelmeyer, Pavlo Holenya, Samuel M. Meier, Suzan Can,
‡
†
∥
‡
,†
̈
Hamed Alborzinia, Julia Schur, Bernhard K. Keppler, Stefan Wolfl, and Ingo Ott*
Institute of Medicinal and Pharmaceutical Chemistry, Technische Universitat
Braunschweig, Germany
†
̈
Braunschweig, Beethovenstraße 55, D-38106
‡
Institute of Pharmacy and Molecular Biotechnology, Ruprecht-Karls-Universitat
Heidelberg, Germany
̈
Heidelberg, Im Neuenheimer Feld 364, D-69120
§
Department of Analytical Chemistry, University of Vienna, Waehringer Straße 38, 1090 Vienna, Austria
∥Institute of Inorganic Chemistry, University of Vienna, Waehringer Straße 42, 1090 Vienna, Austria
*
S Supporting Information
ABSTRACT: Organometallics with N-heterocyclic carbene (NHC) ligands have triggered major interest in inorganic medicinal
chemistry. Complexes of the type Rh(I)(NHC)(COD)X (where X is Cl or I, COD is cyclooctadiene, and NHC is a
dimethylbenzimidazolylidene) represent a promising type of new metallodrugs that have been explored by advanced biomedical
methods only recently. In this work, we have synthesized and characterized several complexes of this type. As observed by mass
spectrometry, these complexes remained stable over at least 3 h in aqueous solution, after which hydrolysis of the halido ligands
occurred and release of the NHC ligand was evident. Effects against mitochondria and general cell tumor metabolism were noted
at higher concentrations, whereas phosphorylation of HSP27, p38, ERK1/2, FAK, and p70S6K was induced substantially already
at lower exposure levels. Regarding the antiproliferative activity in tumor cells, a clear preference for iodido over chlorido
secondary ligands was noted, as well as effects of the substituents of the NHC ligand.
INTRODUCTION
■
Organometallic complexes have been emerging as promising
new metallodrug candidates and hold great promise in
particular for cancer chemotherapy.
1
−3
Among the most
frequently studied types, N-heterocyclic carbene (NHC)
complexes with gold and silver as central atoms have
demonstrated fascinating biochemical properties as enzyme
inhibitors and antimitochondrial agents, suggesting possible
Figure 1. Examples of previously biologically studied rhodium(I)
NHC complexes.
8,9,16,17
4
−6
applications in the therapy of tumors or infectious diseases.
complexes with square-planar geometries. Accordingly, some
analogies with the clinically used platinum drugs are evident,
and in fact, interaction with DNA was found to be a likely mode
7
−9
Early reports
demonstrate that various metals other than
silver or gold can be successfully used to generate biologically
active metal NHC complexes; however, only during the past
few years has this particular field of bioorganometallic
chemistry been gradually extended to different metals including
16
of drug action also for Rh(I) NHC complexes.
Rhodium(III) complexes with ligands other than NHCs
e.g., polypyridyls, staurosporine) have been studied by
(
18,19
20
23
1
0−12
13
14,15
15−17
Sheldrick and co-workers,
Meggers and co-workers,
platinum,
ruthenium, iridium,
and rhodium.
21,22
Leung and Ma and co-workers,
Che and co-workers,
Although Rh(I) NHC complexes were among the first metal
24,25
and others.
Depending on the coordinated ligands for these
NHC complexes to be synthesized, with biological (anti-
8
complexes, several targets were confirmed, including protein
kinases, other enzymes (e.g., the ubiquitin−proteasome
system), and DNA.
bacterial) activity reported in 1996, it was not until 2013 that
their promising potential as anticancer metallodrugs was
described in more detail in three further reports by Simpson
1
5
16
17
et al., McAlpine and co-workers, and us (see Figure 1 for
examples). The use of the Rh(I) center is of special interest
because Rh(I) is isoelectronic with Pt(II), and both ions form
Received: July 23, 2015
Published: November 23, 2015
©
2015 American Chemical Society
9591
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis Procedure for the (COD)Rhodium(I) Derivatives
a
2 2
(
i) Formic acid, reflux, 12 h; (ii) alkylhalide, K CO , CH CN, reflux, 6 h; (iii) Ag O in CH Cl , 4 h; (iv) bis[halido(η ,η -cycloocta-1,5-
2 3 3 2 2 2
diene)rhodium(I)], 2 h.
In this article, we report on interesting chemical-biological
properties that we have observed using Rh(I) complexes 1c/d−
4
c/d of the type Rh(I)(NHC)(COD)X (where X is Cl or I,
COD is cyclooctadiene, and NHC is a dimethylbenzimidazo-
lylidene; see Scheme 1). The dimethylimidazolylidene NHC
fragment was identified as a useful organometallic ligand for
obtaining cytotoxic complexes of this general type in our recent
1
7
report. Initially, the effects of model complexes containing a
-nitrobenzimidazole-based NHC ligand and chlorido or iodido
5
secondary ligands (complexes 2c and 2d) were evaluated. In
these compounds, the electron-withdrawing nitro group
reduces the donor capacity of the NHC ligand, and this
should, in turn, facilitate ligand-exchange interactions with
biomolecular targets. Compounds 2c and 2d can thus be
regarded as model compounds with higher chemical reactivities
and were therefore used to investigate important effects of the
compound type. In a further step, the effects of different
substituents on the 5-position of the NHC ligand on tumor cell
proliferation inhibition were evaluated (1c/d, 3c/d, and 4c/d).
1
Figure 2. H NMR spectra of (top) 4c (X = Cl) and (bottom) 4d (X
= I) in CDCl . (a) N−CH , (b) O−CH , (c) trans CH of COD, (d)
cis CH of COD, (e ) pseudoequatorial CH of COD in cis position,
3
3
3
A
2
(
e ) pseudoequatorial CH of COD in trans position, (f) pseudoaxial
2
CHEMISTRY
B
2
■
CH of COD.
Complexes 1c/d−4c/d were obtained starting from the
respective substituted 1,2-diaminobenzenes or benzimidazoles
(
Scheme 1). Ring closure of the diaminobenzenes with formic
ligand, resulting in a greater steric hindrance toward the CH2
protons in proximity, and this difference in the electronic
surroundings of the protons presumably led to the splitting of
the pseudoequatorial (e and e ) signals. The same effect was
acid afforded the benzimidazoles 1a−3a. Benzimidazole 4a was
commercially available. Next, benzimidazolium cations 1b−4b
were formed by alkylation using methyl iodide in the presence
of K CO . Finally, complexes 1c/d−4c/d were conveniently
obtained by activation of the respective benzimidazolium
cations with Ag O and subsequent transmetalation using
bis[halido(η ,η -cycloocta-1,5-diene)rhodium(I)]. The target
compounds were purified by filtration over Celite and
recrystallized from a mixture of CH Cl and n-hexane. Nuclear
magnetic resonance (NMR) and mass spectrometry (MS)
spectra confirmed the suggested structures, and elemental
analyses indicated their high purities.
A
B
observed with the couples 1c/d−3c/d.
2
3
MASS SPECTROMETRY: HYDROLYTIC STABILITY
2
■
2
2
Electrospray ionization ion-trap mass spectrometry (ESI-IT
MS) was used to study the stability and reactivity of 2c and 2d.
For this purpose, the compounds were dissolved in dimethyl
sulfoxide (DMSO) or dimethylformamide (DMF) (10 mM
concentration), diluted to 100 μM with water, and incubated at
37 °C in the dark, and aliquots withdrawn after different
incubation periods were investigated.
2
2
The target complexes give well-resolved NMR spectra, as
discussed here for 4c and 4d as examples (see Figure 2). In the
Compound 2c is stable for 3 h, irrespective of whether the
solvent is DMSO and DMF, which suggests that these solvents
do not influence the stability of 2c. After 6 and 24 h, the mass
spectra in positive-ion mode of 2c dissolved in DMSO and
DMF were very similar. As evidenced by the mass signal
1
H NMR spectra of 4c and 4d, the shifts of the signals of the
respective methoxy groups (b in Figure 2) show no difference,
whereas the N−CH protons (a in Figure 2) of the iodido
3
derivative 4d are shifted slightly upfield compared to 4c. This
phenomenon can be explained by the difference in the
electronic densities of the halido ligands. The cis- (c in Figure
+
corresponding to [LRh(COD)(OH) − H] , where L is the
nitro-NHC ligand and COD is cyclooctadiene, the compound
hydrolyzes the Rh−Cl bond and forms either aqua or
hydroxido complexes. Loss of the NHC ligand was observed
and seemed to become more pronounced at longer incubation
times, whereas the COD ligand remained coordinated. The
2
) and trans- (d in Figure 2) positioned olefinic signals (cis/
trans relative to X) of the COD moiety are shifted slightly
downfield in the iodido derivative 4d. However, the most
obvious effect observed in the spectra of these derivatives is the
splitting of some of the CH signals (e in Figure 2) in complex
protonated NHC ligand was detected at m/z 192.05 (mtheor =
192.08). Moreover, nonselective solvent adducts with the
2
4d. The iodido ligand has a larger volume than the chlorido
9
592
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
+
organometallic compound were observed for both DMF and
DMSO. Increasing the dry temperature in the ESI source from
(DMSO) − H] were observed in positive-ion mode.
Generally, the iodido analogue 2d undergoes very similar
hydrolysis processes, giving mass signals virtually identical to
those of 2c (Scheme 2). The compound hydrolyzes slightly
faster and forms aqua or hydroxido complexes again
1
80 to 300 °C did not reduce either the DMSO or DMF
adducts, but led to enhanced loss of the NHC ligand.
Nonselective solvent adducts are indicated by broad signals in
the mass spectrum compared to the sharp coordinated adducts;
for example, Figure 3 shows the broad nonselective solvent
+
corresponding to [LRh(COD)(OH) − H] . However, release
of the NHC ligands seems slightly more pronounced for 2d.
The mass spectra of both the DMSO and DMF solutions in
negative-ion mode revealed two major signals at m/z 480.82
and 605.05. MS/MS experiments of these mass signals did not
reveal any additional information on their identity. Further-
−
−
more, mass signals corresponding to I and I3 were detected,
supporting the hypothesis of the hydrolysis pathways.
Of note is the mass signal at m/z 509.96, which was found in
the positive-ion mass spectra of 2c and 2d with DMSO after >6
h of incubation. The identity of the parent mass signal is
unclear, but MS/MS experiments revealed a fragment
+
corresponding to [LRh(OH)(DMSO) − H + Fu] , where
2
Fu is an unknown neutral fragment with m/z 44. This mass
signal increased over time for both compounds, suggesting that
COD release might also represent a possible reaction pathway.
Further details on MS measurements are provided in the
Supporting Information.
EFFECTS ON MITOCHONDRIAL AND CELL
METABOLISM
In previous studies on metal NHC complexes, we observed
■
Figure 3. Sections of the ESI-IT mass spectra in positive-ion mode of
2
c showing the different types of solvent adducts with (A) DMSO and
strong effects against mitochondria as well as tumor cell
(
B) DMF. Broad mass signals indicate nonselective adducts between
13,27,28
metabolism.
Analogous experiments with 2c and 2d
the solvent and metal, and sharp peaks indicate solvent coordination to
the metal.
showed some comparable trends (e.g., a decrease of the
mitochondrial membrane potential with both compounds,
inhibition of cell respiration with 2d); however, the effects were
less evident and appeared significant only at much higher
dosages (e.g., 50 μM). A more detailed discussion of these
results is presented in the Supporting Information.
adduct of hydrolyzed 2c at m/z 480.06 and the respective
complex giving a sharp mass signal at m/z 402.06. The
negative-ion mode spectra of 2c dissolved in DMSO and DMF
feature a single mass signal that increases over time and
corresponds to a dimeric Rh complex that has undergone
−
ELISA MICROARRAYS
carbene release, namely, [Rh (COD) (μ-Cl) (OH) − 2H]
■
2
2
3
2
(Figure S1B, Supporting Information). Rhodium has a strong
To analyze the effects of complexes 2c and 2d on cell signaling,
enzyme-linked immunosorbant assay (ELISA) microarrays
were applied to detect the absolute levels of a selection of
key signaling proteins in their phosphorylated states. To obtain
time-resolved images of signaling modulation, HT-29 cells were
affinity for chloride, and a similar type of Rh-chlorido dimer
was reported that was also accompanied by ligand release.
26
Compound 2d is also stable for 3 h in DMF and DMSO,
after which increases in free carbene and [Rh(COD)(OH)-
−
−
Scheme 2. Proposed Hydrolysis Pathway of 2c (X = Cl ) and 2d (X = I ) in Aqueous Solution
9
593
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
S78/S82
, (B) phospho-p38T180/Y182 (no legend), (C) phospho-ERK1T202/Y204, (D)
, (G) phospho-WNK1 , (H) phospho-GSK-3β .
Figure 4. ELISA microarrays with 2c and 2d: (A) phospho-HSP27
phospho-ERK2
T185/Y187
Y397
T421/S424
T60
S9
, (E) phospho-FAK , (F) phospho-p70S6
incubated in time-course experiments (0−24 h) with two
concentrations of each compound, namely, 2.5 and 5.0 μM
phospho-ERK1 (T202/Y204), phospho-ERK2 (T185/Y187),
and phospho-p38α (T180/Y182) increased 2−10-fold upon
treatment with 5.0 μM 2c compared to mock-treated (0.1% v/v
DMF) samples. The same concentration of 2d showed slightly
weaker effects on MAPK activation. ERK1/2 and p38 are
generally known to be activated in response to the intracellular
(Figure 4), and cell lysates collected at indicated time points
were incubated with the microarrays.
We observed a significant time-dependent activation of
mitogen-activated protein kinases (MAPKs). The levels of
9
594
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
redox state and oxidative stress and potentially contribute to
influencing cell survival or cell death.
MDA-MB-231 cells over HT-29 cells. Established anticancer
drugs such as cisplatin and 5-fluorouracil trigger comparable
cytotoxic effects in the range of approximately 1−10 μM in this
29
Our analysis also revealed a strong increase of chaperone
HSP27 phosphorylation. This protein mediates the protection
properties of cells in response to cytotoxic stress (e.g., loss of
protein functionality upon binding of organometallic frag-
35,36
assay.
Regarding the substituents on the NHC ligand, the
methyl group of 3c/d triggered the lowest IC50 values in both
cell lines, and regarding the secondary halido ligands, the iodido
derivatives 1d−4d were in most cases more active than the
respective chlorido complexes 1c−4c.
30
ments), growth arrest, or receptor-mediated apoptosis.
Interestingly, phospho-FAK (Y397), known to regulate focal
31
contacts and cell detachment, exhibited a transient activation
profile with both compounds: 5 μM 2c and 2d induced a slight
increase in its phosphorylation during 2 h of treatment followed
by a sustained decrease. The levels of phospho-p70S6 kinase, a
key downstream effector of mTORC1, were also increased
upon treatment, indicating changes in cellular protein syn-
CONCLUSIONS
■
Complexes of the type Rh(I)(NHC)(COD)X (where X = Cl
or I) can be conveniently prepared following an established
method that is based on the reaction of the precursor
3
2
benzimidazolium cations with Ag O and a subsequent trans-
thesis. Several signaling proteins, among them the key protein
of PI3 kinase signaling, Akt1; the important transcription factor
CREB; and the DNA damage checkpoint Chk2 showed only
minor or insignificant changes in phosphorylation (see
Supporting Information). Finally, we observed a decrease in
phosphorylated ERK5 signaling upstream modulator WNK1,
2
2
2
metalation reaction using bis[halido(η ,η -cycloocta-1,5-diene)-
1
rhodium(I)]. H NMR spectra confirmed differing electronic
effects of the chlorido and iodido ligands, which might indicate
different kinetic reactivities. The two organometallic Rh
compounds 2c and 2d were stable in aqueous solution
containing 1% DMF or 1% DMSO for at least 3 h. After 6 h,
however, the nitro-NHC ligand was released to a significant
degree and might provide a route of decomposition. In
contrast, the COD ligand remained largely coordinated
throughout the incubation period, although the mass signal at
m/z 509.97 was indicative of some COD release. The mass
spectra of the DMSO and DMF solutions were very similar,
featuring several selective and nonselective solvent adducts with
the organometallics. Compound 2c eventually forms a dimeric
33
implicated in antiapoptotic signaling. GSK-3β was also
dephosphorylated (activated) upon treatment. This enzyme is
a key regulator of numerous signaling pathways, including
cellular responses to Wnt, receptor tyrosine kinases, and G
protein-coupled receptors and is involved glycogen metabolism
34
and cell cycle regulation and proliferation.
Table 1. Antiproliferative Effects against HT-29 and MDA-
MB-231 Cells
compound corresponding to [Rh (COD) (μ-Cl) (OH)
−
2
2
3
2
IC50 HT-29 (μM)
IC50 MDA-MB-231 (μM)
−
2H] , whereas unambiguous decomposition products of 2d are
still elusive.
1
2
3
4
1
2
3
4
1
1
2
2
3
3
4
4
a
a
a
a
b
b
b
b
c
>100
>100
>100
>100
Of note, in a previous study, we observed rapid cellular
uptake of the structurally related complex Rh(I)(NHC)(COD)
>100
>100
17
>100
>100
Cl (see Figure 1) within 1−4 h. Accordingly, it can be
speculated that the complexes reach the cellular environment in
an intact form and that the mentioned ligand-exchange
processes occur mainly inside the cells, leading to bioactive
metabolites. Importantly, the observed products of hydrolysis
are of a cationic nature, and this indicates some analogies with
the biochemistry of cisplatin, which forms cationic aqua
complexes upon intracellular hydrolysis that represent the
>100
>100
>100
>100
>100
>100
>100
>100
12.1 ± 1.5
5.1 ± 0.2
10.2 ± 2.2
8.6 ± 0.6
5.2 ± 1.6
1.5 ± 0.3
9.5 ± 0.4
6.6 ± 0.6
9.0 ± 0.9
2.7 ± 1.2
5.2 ± 0.1
6.4 ± 1.2
2.4 ± 0.6
1.5 ± 0.1
4.1 ± 0.6
3.5 ± 0.3
d
c
d
c
37
species attacking DNA. Also regarding cellular metabolism,
interactions with molecular targets have to be considered (e.g.,
glutathione or other cellular thiols).
d
c
The above-noted differences between the iodido and
chlorido ligands also were reflected in certain differences in
their biological activities. Whereas the chlorido derivative 2c
was a stronger uncoupling-like agent in mitochondria, the
iodido complex 2d triggered stronger activity against general
cellular metabolism (impedance, respiration, acidification; see
the Supporting Information for details). However, the observed
activities required rather high concentrations and can therefore
be considered as secondary effects.
d
ANTIPROLIFERATIVE EFFECTS IN TUMOR CELLS
■
To evaluate the effects on proliferation inhibition of different
substituents on the NHC ligand, we screened complexes
containing different donor/acceptor groups against tumor cell
lines. Complexes 1c/d−4c/d along with their respective
benzimidazole and benzimidazolium iodide precursors were
investigated on two tumor cell lines (namely, HT-29 colon
carcinoma and MDA-MB-231 human breast adenocarcinoma).
The precursor [Rh(I)Cl(COD)]2 without NHC ligand is
Clear effects at lower concentrations were observed on
cellular signaling determined by ELISA microarrays that had
previously been developed to study the activation of major
pathways controlling stress response and cellular prolifer-
17
38,39
inactive in terms of cytotoxicity, as reported earlier. Whereas
the metal free benzimidazoles and benzimidazolium cations 1a/
b−4a/b were not active, the rhodium complexes 1c/d−4c/d
exhibited half-maximal inhibitory concentration (IC ) values in
ation.
In these experiments, the two complexes triggered
comparably strong phosphorylations of HSP27, and MAPKs
such as p38 and ERK1/2 were also significantly activated,
reflecting a general cellular stress response. In the case of
MAPK phosphorylation, the chlorido complex 2c showed
5
0
the low micromolar range (1.5−12.1 μM) with preference for
9
595
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
higher efficiency. Other important signaling targets that were
modulated included FAK and p70S6.
EXPERIMENTAL SECTION
■
General. All reagents and solvents were used as received from
Sigma, Aldrich, or Acros. 5-Methoxybenzimidazole (4a) was obtained
The observed phosphorylation profiles differ significantly
from results obtained for cisplatin and camptothecin, which
directly target DNA or replication. In comparable assays,
although under slightly different conditions and with another
cancer cell line in the case of cisplatin, these compounds did
1
13
from Acros. H NMR and C NMR spectra were recorded on a
Bruker DRX-400 AS NMR System; MS spectra were recorded on a
FinniganMAT4515. The purities of the target compounds (>95%)
were confirmed by elemental analysis (Flash EA112, Thermo Quest
Italia). For all compounds undergoing biological evaluation, the
experimental values differed by less than 0.5% from the calculated
40
not lead to ERK1/2 phosphorylation. Within 8−11 h,
cisplatin also did not affect Akt1 and GSK-3β phosphorylation
40
ones.
levels but showed a decrease after 24 h. Complexes 2c and 2d
initially decreased GSK-3β phosphorylation, but levels
increased after longer incubation. Akt1 was not strongly
influenced by 2c and 2d (small decrease of phosphorylation).
These results indicate that the mode of action of rhodium(I)
NHC complexes differs significantly from the activities of
cisplatin or camptothecin. This is further confirmed by the only
weak activation of Chk2 that is activated in response to DNA
44
5
-Chlorobenzimidazole (1a). 4-Chloro-1,2-phenylendiamine
(
1
1.891 g, 13.26 mmol) and 40 mL of formic acid were added to a
00 mL flask and refluxed at 110 °C for 4 h, giving a deep blue
solution. The mixture was cooled to 0 °C, and the product was
precipitated by neutralization with concentrated ammonia. 1a was
filtered off, washed twice with water, and dried at 50 °C. Yield: 1.911 g
12.53 mmol, 95%), purple powder. H NMR (CDCl , ppm): 10.43
1
(
(
3
4
s, 1H, NH), 8.20 (s, 1H, ArH2), 7.65 (d, 1H, J = 2.0 Hz, ArH4), 7.55
3
4
3
38
(d, 1H, J = 8.8 Hz, ArH7), 7.27 (dd, 1H, J = 2.0 Hz, J = 8.8 Hz,
damage and replication stress. Regarding a comparison with
recently studied gold species, complexes 2c and 2d showed a
pattern similar to that of previously studied phosphane-
containing complexes (e.g., auranofin) that also triggered
ERK, p38, and HSP27 activation but differing from that of a
biscarbene gold complex, which, again, did not activate ERK
13
ArH6). C NMR (CDCl ): 141.3 (ArC-2), 138.2 (ArC-Cl), 136.3
3
(
7
ArC-7a), 128.8 (ArC-3a), 123.7 (ArC-6), 116.4 (ArC-4), 115.4 (ArC-
). Elemental analysis for C H ClN (% calcd/found): C (55.10/
7
5
2
54.20), H (3.30/3.30), N (18.36/17.87).
44
5-Nitrobenzimidazole (2a). 4-Nitro-1,2-phenylendiamine (1.784
g, 11.65 mmol) and 40 mL of formic acid were added to a 100 mL
flask and refluxed at 110 °C for 4 h, giving a green solution. The
mixture was cooled to 0 °C, and the product was precipitated by
neutralization with concentrated ammonia. 2a was filtered off, washed
4
1,42
phosphorylation.
Screening of an extended series of Rh(I)(NHC)(COD)X
complexes for their antiproliferative effects against two cancer
lines showed, in general, a certain preference for iodido over
chlorido ligands. However, the couple 2c/2d represented an
exception where no clear trend could be noted in this respect.
Advantages of switching from chlorido to iodido complexes
were previously reported by Sadler and co-workers for
organometallic complexes of the type [M(p-cymene)(azo/
twice with water, and dried at 50 °C. Yield: 1.621 g (9.94 mmol, 85%),
1
light green powder. H NMR (DMSO-d , ppm): 13.10 (s, 1H, NH),
6
5
4
8
.56 (s, 1H, ArH2), 8.52 (dd, 1H, J = 0.5 Hz, J = 2.3 Hz, ArH4), 8.15
4
3
5
3
(dd, 1H, J = 2.3 Hz, J = 8.8 Hz, ArH6), 7,77 (dd, 1H, J = 0.5 Hz, J
= 8.8 Hz, ArH7). 13C NMR (DMSO-d ): 146.8 (ArC-2), 142.6
6
(ArCC-NO
), 117.6 (ArC-6), 114.9 (ArC-4), 112.8 (ArC-7).
2
Elemental analysis for C H O N (% calcd/found): C (51.54/
7
5
2
3
+
51.48), H (3.09/2.97), N (25.76/25.65).
imino-pyridine)X] , where M was Ru or Os and X was Cl or
45
4
3
5-Methylbenzimidazole (3a). 4-Methyl-1,2-phenylendiamine
I. In that case, the iodido complexes were also more potent
and differed in other anticancer-related properties such as p53
dependency for activity.
(
1
0.579 g, 4.74 mmol) and 20 mL of formic acid were added to a
00 mL flask and refluxed at 110 °C for 4 h, giving a red-brown
solution. The mixture was cooled to 0 °C and neutralized with
concentrated ammonia, and the solvent was removed under reduced
pressure. 3a was extracted with dichloromethane, which was removed
by reduced pressure. The product was highly hydroscopic brown oil.
Yield: 0.510 g (3.86 mmol, 81%), brown oil. H NMR (CDCl , ppm):
2.73 (s, 1H, NH), 8.26 (s, 1H, ArH2), 7.54 (d, 1H, J = 8.3 Hz,
Further conclusions regarding structure−activity relation-
ships for cytotoxicity indicated a preference for the methyl
group (see results for 3c/d in Table 1), which exhibits a
positive inductive effect that might enhance the donor capacity
of the NHC ligand, or the unsubstituted NHC ligand, which
1
3
3
1
3
17
ArH7), 7.44 (s, 1H, ArH4), 7.13 (d, 1H, J = 8.3, ArH6), 2.4 (s, 3H,
afforded higher cytotoxicity in our previous study. Taken
together, future lead-optimization strategies for Rh(I) NHC
complexes should further address the coordinated halide
ligands as well as the fine-tuning of donor properties of the
NHC ligand system. Regarding our previous report, solubility
plays a deciding role, and therefore, rather low-lipophilicity
NHC ligands should be preferred in future studies.
Replacement of the chlorido ligand with CO has already
been investigated, but it turned out to be problematic in terms
−
(
6
(
CH3). 13C NMR (CDCl ): 139.7 (ArC-2), 135.6 (ArC−CH ), 134.1
3
3
ArC-3a), 133.9 (ArC-7a), 125.4 (ArC-7), 114.8 (ArC-4), 114.5 (ArC-
), 21.5 (−CH ). Elemental analysis for C H N (% calcd/found): C
3
8
8
2
72.70/69.78), H (6.10/5.95), N (21.20/19.84).
General Procedure for Synthesis of the Benzimidazolium
Iodides 1b−4b. One equivalent of substituted benzimidazole, 1 equiv
of K CO , and an excess of methyl iodide were heated under reflux in
2
3
acetonitrile/toluene (1:1) for 12 h. The solvent was removed under
reduced pressure, and the resultant solid was resuspended in
dichloromethane and filtered to remove the formed potassium iodide
along with the remaining K CO . The solvent of the filtrate was
17
of lipophilicity and stability in solution.
2
3
In summary, our study shows that the biological properties of
rhodium(I) NHC complexes can be modulated by changing the
substituents on the NHC ligands and the nature of the
secondary halide ligand. Regarding their mode of drug action,
effects on cellular signaling of certain targets such as p38,
ERK1, and ERK2 have to be taken into account.
removed under reduced pressure, and the product resuspended in
tetrahydrofuran to remove the excess methyl iodide and unreacted
substituted benzimidazole.
5
-Chloro-1,3-dimethylbenzimidazolium Iodide (1b). 1a (0.391 g,
2
1
.56 mmol), K CO (0.354 g, 2.56 mmol) and methyl iodide (0.7 mL,
0.25 mmol) were dissolved in 20 mL of acetonitrile/toluene (1:1)
2
3
8
,9,15−17
and refluxed for 12 h. Yield: 0.354 g (1.95 mmol, 76%), purple
Taken together with previous reports,
this study
1
4
powder. H NMR (CDCl , ppm): 9.69 (s, 1H, ArH2), 8.28 (d, 1H, J
=
3
shows that the Rh(I)(NHC)(COD) fragment might represent
a useful organometallic pharmacophore for the design of novel
anticancer and antibacterial agents that interfere with MAPK
signaling and other yet-to-be-identified targets.
3
4
2.0 Hz, ArH4), 8.06 (d, 1H, J = 8.8 Hz, ArH7), 7.78 (dd, 1H, J =
3
2
.0 Hz, J = 8.8 Hz, ArH6), 4.08 (s, 3H, N−CH ), 3.98 (s, 3H, N−
3
13
CH ). C NMR (CDCl ): 156.2 (ArC-Cl), 144.3 (ArC-2), 132.5
3
3
(ArC-7a), 131.1 (ArC-3a), 126.7 (ArC-6), 115.1 (ArC-4), 113.7 (ArC-
9
596
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
2
2
7
), 34.3 (N−CH ), 34.1 (N−CH ). Elemental analysis for
(0.042 g, 0.18 mmol), bis[iodido(η ,η -cycloocta-1,5-diene)rhodium-
3
3
C H ClN I (% calcd/found): C (35.03/36.45), H (3.27/3.39), N
(I)] (0.085 g, 0.12 mmol) , 2 h. Yield: 0.044 g (0.09 mmol, 28%), light
brown powder. H NMR (CDCl , ppm): 7.28 (dd, 1H, J = 0.6 Hz, J
= 1.7 Hz, ArH4), 7.20 (dd, 1H, J = 1.7 Hz, J = 8.5, ArH6), 7.18 (dd,
9
10
2
1
5
4
(9.08/8.50)
3
4
3
1
,3-Dimethyl-5-nitrobenzimidazolium Iodide (2b). 2a (0.313 g,
5
3
1
7
.92 mmol), K CO (0.300 g, 1.92 mmol), and methyl iodide (0.5 mL,
1H, J = 0.6 Hz, J = 8.5, ArH7), 5.37 (s, 2H, CH−COD), 4.18 (s, 3H,
2
3
.67 mmol) were dissolved in 20 mL of acetonitrile/toluene (1:1) and
N−CH ), 4.17 (s, 3H, N−CH ), 3.52 (s, 2H, CH−COD), 2.38 (m,
3
3
refluxed for 12 h. Yield: 0.338 g (1.75 mmol, 91%), light brown
powder. H NMR (DMSO-d , ppm): 9.94 (s, 1H, ArH2), 9.08 (dd,
1
4H, CH −COD), 2.05 (m, 2H, CH −COD), 1.88 (m, 2H, CH −
2
2
2
1
13
1
COD). C NMR (CDCl ): 198.9 (d, J = 49.5 Hz, NHC), 136.1
6
3
5
4
4
3
H, J = 0.5 Hz, J = 2.3 Hz, ArH4), 8.57 (dd, 1H, J = 2.3 Hz, J = 9.0
(ArC-Cl), 134.2 (ArC-7a), 128.5 (ArC-3a), 122.5 (ArC-4), 109.8
5
3
1
1
Hz, ArH6), 8.27 (dd, 1H, J = 0.5 Hz, J = 9.0 Hz, ArH7), 4.19 (s, 3H,
(ArC-6), 109.5 (ArC-7), 98.6 (d, J = 5.2 Hz, CH−COD), 98.5 (d, J
13
1
1
N−CH ), 4.15 (s, 3H, N−CH ). C NMR (DMSO-d ): 147.4 (ArC-
= 5.2 Hz, CH−COD), 71.9 (d, J = 14.0 Hz, CH−COD), 71.8 (d, J =
14.0 Hz, CH−COD), 34.9 (N−CH ), 34.8 (N−CH ), 32.3 (CH −
3
3
6
2), 145.5 (ArC-NO ), 135.3 (ArC-7a), 131.5 (ArC-3a), 121.4 (ArC-6),
2
3
3
2
1
14.8 (ArC-4), 110.8 (ArC-7), 33.8 (2× N−CH ). Elemental analysis
COD), 32.3 (CH −COD), 29.5 (CH −COD), 29.5 (CH −COD);
3
2
2
2
+
for C H O N I (% calcd/found): C (33.88/33.97), H (3.16/3.27), N
MS(EI): 518 (M ). Elemental analysis for C H N ClIRh (% calcd/
9
10
2
3
17 21 2
(
13.17/11.94).
,3,5-Trimethylbenzimidazolium Iodide (3b). 3a (0.510 g, 3.86
mmol), K CO (0.540 g, 3.86 mmol), and methyl iodide (1.0 mL,
found): C (39.37/40.67), H (4.08/4.05), N (5.40/5.25).
4
6
2
2
1
Chlorido(η ,η -cycloocta-1,5-diene)(5-nitro-1,3-dimethylbenzimi-
dazol-2-ylidene)rhodium(I) (2c). 2b (0.102 g, 0.32 mmol), Ag
O
2
3
2
2
2
1
5.44 mmol) were dissolved in 20 mL of acetonitrile/toluene (1:1)
(0.044 g, 0.19 mmol), bis[chlorido(η ,η -cycloocta-1,5-diene)rhodium-
and refluxed for 12 h. Yield: 0.611 g (2.12 mmol, 54.9%), white
powder. H NMR (CDCl , ppm): 11.03 (s, 1H, ArH2), 7.57 (d, 1H, J
=
(
(I)] (0.089 g, 0.16 mmol), 2 h. Yield: 0.117 g (0.27 mmol, 84%),
1
3
1
4
3
yellow powder. H NMR (CDCl
, ppm): 8.24 (dd, 1H, J = 2.1 Hz, J
3
3
4
3
4
3
8.6 Hz, ArH7), 7.51 (dd, 1H, J = 1.0 Hz, J = 8.6 Hz, ArH6), 7.48
= 8.6 Hz, ArH6), 8.22 (d, 1H, J = 2.1 Hz, ArH4), 7.35 (d, 1H, J = 8.6
4
d, 1H, J = 1.0 Hz, ArH4), 4.24 (s, 3H, N−CH ), 4.23 (s, 3H, N−
Hz, ArH7), 5.24 (s, 2H, CH−COD); 4.40 (s, 3H, N−CH
3H, N−CH ), 3.40 (s, 2H, CH−COD), 2.50 (m, 4H, CH
2.09 (m, 4H, CH
Hz, NHC), 143.4 (ArC-NO
(ArC-6), 109.0 (ArC-4), 105.5 (ArC-7), 101.9 (d, J = 6.5 Hz, CH−
3
), 4.39 (s,
−COD),
3
13
CH ), 2.61 (s, 3H, −CH ). C NMR (CDCl ): 142.3 (ArC-2), 138.5
3
2
3
3
3
1
3
1
(
(
(
(
ArC−CH ), 132.1 (ArC-3a), 130.0 (ArC-7a), 129.1 (ArC-7), 112.2
2
−COD). C NMR (CDCl
): 204.5 (d, J = 51.0
3
3
ArC-6), 112.2 (ArC-4), 33.8 (N−CH ), 33.7 (N−CH ), 21.9
2
), 138.7 (ArC-7a), 134.9 (ArC-3a), 118.7
3
3
1
−CH ). Elemental analysis for C H N I (% calcd/found): C
3
10 13
2
1
1
41.69/41.46), H (4.55/4.50), N (9.72/9.72).
-Methoxy-1,3-dimethylbenzimidazolium Iodide (4b). 5-Methox-
ybenzimidazole (0.706 g, 4.77 mmol), K CO (0.700 g, 4.77 mmol),
COD), 101.8 (d, J = 6.5 Hz, CH−COD), 68.9 (d, J = 14.2 Hz, CH−
1
5
COD), 68.8 (d, J = 14.2 Hz, CH−COD), 35.2 (N−CH
CH ), 33.0 (CH −COD), 33.0 (CH −COD), 28.8 (CH
28.8 (CH −COD); MS(EI): 437 (M ). Elemental analysis for
ClRh (% calcd/found): C (46.65/46.85), H (4.84/
5.04), N (9.60/8.05).
3
). 35.2 (N−
−COD),
2
3
3
2
2
2
+
and methyl iodide (1.2 mL, 19.06 mmol) were dissolved in 20 mL of
acetonitrile/toluene (1:1) and refluxed for 12 h. Yield: 0.944 g (3.10
mmol, 65%), brown powder. H NMR (CDCl , ppm): 10.73 (s, 1H,
ArH2), 7.58 (dd, 1H, J = 1.6 Hz, J = 7.8 Hz, ArH7), 7.21 (dd, 1H, J
=
3
2
C H O N
17 21 2 3
1
3
4
3
4
2
2
(η ,η -Cycloocta-1,5-diene)iodido(5-nitro-1,3-dimethylbenzimi-
dazol-2-ylidene)rhodium(I) (2d). 2b (0.091 g, 0.29 mmol), Ag O
2
0.025 g, 0.10 mmol), bis[iodido(η ,η -cycloocta-1,5-diene)rhodium-
3
4
2.3 Hz, J = 7.8 Hz, ArH6), 7.16 (d, 1H, J = 2.3 Hz, ArH4), 4.22 (s,
2
2
13
(
H, N−CH ), 4.19 (s, 3H, N−CH ), 3.95 (s, 3H, −OCH ). C NMR
3
3
3
(
(
(
CDCl ): 159.5 (ArC-OCH ), 153.8 (ArC-2), 141.6 (ArC-3a), 132.8
(I)] (0.050 g, 0.07 mmol), 2 h. Yield: 0.044 g (0.011 mmol, 38%),
3
3
1
5
ArC-7a), 116.9 (ArC-7), 113.3 (ArC-4), 95.5 (ArC-6), 56.8
light brown powder. H NMR (CDCl
3
, ppm): 8.23 (dd, 1H, J = 0.6
4
4
3
−OCH ), 33.4 (N−CH ), 33.1 (N−CH ). Elemental analysis for
Hz, J = 2.1 Hz, ArH4), 8.21 (dd, 1H, J = 2.1 Hz, J = 8.5 Hz, ArH6),
3
3
3
5
3
C H ON I (% calcd/found): C (39.49/40.26), H (4.31/4.31), N
7.36 (dd, 1H, J = 0.6 Hz, J = 8.5 Hz, ArH7), 5.42 (s, 2H, CH−
COD), 4.29 (s, 3H, N−CH ), 4.28 (s, 3H, N−CH ), 3.56 (s, 2H,
CH−COD), 2.42 (m, 4H, CH −COD), 2.14 (m, 2H, CH −COD),
−COD). C NMR (CDCl
Hz, NHC), 143.2 (ArC-NO
10
13
2
(
9.21/9.04).
3
3
General Procedure for Synthesis of Rhodium NHC Com-
2
2
13
1
plexes 1c/d−4c/d. One equivalent of the respective benzimidazo-
1.91 (m, 2H, CH
2
): 205.3 (d, J = 48.6
3
lium iodide 1b−4b and 0.5 equiv of Ag O were added to a dried
2
), 138.2 (ArC-7a), 135.1 (ArC-3a), 118.7
2
1
Schlenk tube. The mixture was back-flashed three times with N , and
(ArC-6), 108.8 (ArC-4), 105.3 (ArC-7), 99.6 (d, J = 6.2 Hz, CH−
1 1
2
then 15 mL of dry CH Cl 2 were added. The flask was closed, and the
mixture was stirred for 4 h in the dark. A solution containing 0.5 equiv
of bis[halido(η ,η -cycloocta-1,5-diene)rhodium(I)] in CH Cl was
added (10 mL), and the solution was stirred for 2 h in the dark. The
obtained suspension was filtered over Celite (281 nm) and
concentrated in a vacuum. The yellow residue was recrystallized
from CH Cl /n-hexane (10 mL/40 mL) at 4 °C.
COD), 99.5 (d, J = 6.2 Hz, CH−COD), 72.3 (d, J = 14.2 Hz, CH−
2
1
COD), 72.2 (d, J = 14.2 Hz, CH−COD), 35.3 (N−CH ), 35.2 (N−
3
2
2
CH
), 32.3 (CH
3
−COD), 32.3 (CH
−COD), 29.4 (CH −COD),
2
2
2
2
2
+
29.4 (CH −COD); MS(EI): 529 (M ). Elemental analysis for
2
C H O N IRh (% calcd/found): C (38.58/39.06), H (4.00/4.10),
17
21
2
3
N (7.94/7.78).
2
2
Chlorido(η ,η -cycloocta-1,5-diene)(1,3,5-trimethylbenzimidazol-
2-ylidene)rhodium(I) (3c). 3b (0.092 g, 0.32 mmol), Ag O (0.053 g,
0.23 mmol), bis[chlorido(η ,η -cycloocta-1,5-diene)rhodium(I)]
2
2
2
2
Chlorido(5-chloro-1,3-dimethylbenzimidazol-2-ylidene)(η ,η -cy-
2
2
2
cloocta-1,5-diene)rhodium(I) (1c). 1b (0.096 g, 0.31 mmol), Ag O
2
2
2
(
0.042 g, 0.18 mmol), bis[chlorido(η ,η -cycloocta-1,5-diene)rhodium-
(0.082 g, 0.15 mmol), 2 h. Yield: 0.082 g (0.20 mmol, 63%), yellow
1
3
(
I)] (0.078 g, 0.14 mmol), 2 h. Yield: 0.054 g (0.13 mmol, 41%),
powder. H NMR (CDCl , ppm): 7.14 (d, 1H, J = 8.2 Hz, ArH7),
3
1
5
5
5
3
yellow powder. H NMR (CDCl , ppm): 7.28 (dd, 1H, J = 0.5 Hz, 4J
=
(
4
2
7.07 (t, 1H, J = 0.8 Hz, ArH4), 7.03 (ddd, 1H, J = 0.8 Hz, J = 8.2
3
4
3
1.7 Hz, ArH4), 7.21 (dd, 1H, J = 1.7 Hz, J = 8.5 Hz, ArH6), 7.18
Hz, ArH6), 5.15 (s, 2H, CH−COD), 4.27 (s, 3H, N−CH ), 4.26 (s,
3
5
3
dd, 1H, J = 0.5 Hz, J = 8.5 Hz, ArH7), 5.17 (s, 2H, CH−COD),
3H, N−CH ), 3.35 (s, 2H, CH−COD), 2.46 (m, 4H, CH −COD),
3
2
13
.29 (s, 3H, N−CH ), 4.28 (s, 3H, N−CH ), 3.36 (s, 2H, CH−COD),
2.46 (s, 3H, CH ), 2.01 (m, 4H, CH −COD). C NMR (CDCl ):
3
3
3
2
3
13
1
.47 (m, 4H, CH −COD), 2.02 (m, 4H, CH −COD). C NMR
195.2 (d, J = 50.6 Hz, ArC-2), 135.6 (ArC−CH3), 133.5 (ArC-3a),
2
2
1
(
CDCl ): 198.3 (d, J = 50.9 Hz, ArC-2), 135.9 (ArC-Cl), 134.0 (ArC-
132.5 (ArC-7a), 123.8 (ArC-7), 109.5 (ArC-4), 108.8 (ArC-6), 100.1
3
1
1
7
1
6
a), 128.7 (ArC-3a), 122.7 (ArC-6), 109.9 (ArC-4), 109.6 (ArC-7),
(d, J = 2.3 Hz, CH−COD), 100.0 (d, J = 2.3 Hz, CH−COD), 68.3
1
1
1
1
00.8 (d, J = 6.6 Hz, CH−COD), 100.7 (d, J = 6.6 Hz, CH−COD),
(d, J = 14.5 Hz, CH−COD), 68.2 (d, J = 14.5 Hz, CH−COD), 34.6
(N−CH3), 34.5 (N−CH3), 32.9 (CH2-COD), 32.9 (CH2-COD),
28.9 (CH −COD), 28.9 (CH −COD), 21.8 (−CH ); MS(EI): 406
1
1
8,5 (d, J = 14.6, CH−COD), 68.4 (d, J = 14.6, CH−COD), 34.8 (s,
N−CH ), 34.8 (s, N−CH ), 32.9 (CH −COD), 32.9 (CH −COD),
2
analysis for C H N2Cl Rh (% calcd/found): C (47.80/47.93), H
3
3
2
2
2
2
3
+
+
8.8 (CH −COD), 28.8 (CH −COD); MS(EI): 426 (M ). Elemental
(M ). Elemental analysis for C H N ClRh (% calcd/found): C
2
2
18 24 2
(53.15/53.11), H (5.95/5.87), N (6.89/6.53).
17
21
2
2
2
(
4.96/5.55), N (6.56/5.05).
5-Chloro-1,3-dimethylbenzimidazol-2-ylidene)(η ,η -cycloocta-
,5-diene)iodidorhodium(I) (1d). 1b (0.093 g, 0.31 mmol), Ag O
(η ,η -Cycloocta-1,5-diene)iodido(1,3,5-trimethylbenzimidazol-2-
2
2
(
ylidene)rhodium(I) (3d). 3b (0.077 g, 0.27 mmol), Ag O (0.029 g,
2
2
2
1
0.13 mmol), bis[iodido(η ,η -cycloocta-1,5-diene)rhodium(I)] (0.098
2
9
597
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
g, 0.13 mmol). Yield: 0.012 g (0.02 mmol, 9%), light brown powder.
added to cells without medium replacement (final DMF concentration
= 0.1% v/v). For mock treatment, cells were incubated with 0.1% (v/
v) DMF only. At indicated time points, cells were washed three times
with ice-cold D-PBS (Invitrogen) and lysed in 100 μL lysis buffer (6 M
urea, 1 mM EDTA, 0.5% Triton X-100, 5 mM NaF, 10 μg/mL
leupeptin, 10 μg/mL pepstatin, 100 μM PMSF, 3 μg/mL aprotinin, 2.5
mM Na P O , 1 mM Na VO in D-PBS). Collected samples were
1
5
3
H NMR (CDCl , ppm): 7.13 (d, 1H, J = 8.2 Hz, ArH7), 7.07 (t, 1H,
3
5
3
J = 0.9 Hz, ArH4), 7.04 (dd, 1H, J = 0.9 Hz, J = 8.2, ArH6), 5.32 (s,
H, CH−COD), 4.16 (s, 6H, N−CH ), 3.51 (s, 2H, CH−COD), 2.46
2
(
1
3
s, 3H, −CH ), 2.37 (m, 4H, CH −COD), 2.03 (m, 2H, CH −COD),
3
2
2
1
3
1
.85 (m, 2H, CH −COD). C NMR (CDCl ): 192.4 (d, J = 53.6
2
3
Hz, NHC), 135.3 (ArC−CH ), 133.6 (ArC-3a), 133.2 (ArC-7a), 124.4
4
2
7
3
4
3
1
centrifuged for 15 min at 4 °C and 13.000 rpm, and supernatants were
(
ArC-7), 109.9 (ArC-4), 109.4 (ArC-6), 91.2 (d, J = 7.2 Hz, CH−
1
1
frozen at −80 °C for further analysis.
COD), 91.1 (d, J = 7.2 Hz, CH−COD), 35.9 (d, J = 15.4 Hz, CH−
1
Total Protein Concentration. Total protein concentration was
determined using the BCA Protein Assay (Pierce Biotechnology) in a
96-well-plate format.
COD), 35.8 (d, J = 15.4 Hz, CH−COD), 33.8 (N−CH ), 33.8 (N−
3
CH ), 30.8 (CH −COD), 30.8 (CH −COD), 30.7 (CH −COD),
3
2
2
2
+
3
0.7 (CH −COD), 21.4 (−CH ); MS(EI): 498 (M ). Elemental
2
3
ELISA Microarray Protocol. Proteins were quantified using
sandwich ELISA microarrays. The microarrays were based on the
ArrayStripTM platform (Alere Technologies GmbH). A detailed
description of the assay protocol, information on the reagents for this
assay, and a list of the currently available targets were previously
analysis for C H N IRh (% calcd/found): C (43.39/43.69), H
1
8
23
2
(
4.86/4.70), N (5.62/4.98).
Chlorido(η ,η -cycloocta-1,5-diene)(1,3-dimethyl-5-methoxy-
2
2
benzimidazol-2-ylidene)rhodium(I) (4c). 4b (0.085 g, 0.28 mmol),
Ag O (0.050 g, 0.21 mmol), bis[chlorido(η ,η -cycloocta-1,5-diene)-
2
2
2
3
9
reported. In brief, cellular samples were diluted 1:6 with dilution
buffer (1 mM EDTA, 0.5% Triton X-100, 5 mM sodium fluoride, 1 M
urea in buffered saline, pH 7.2) and incubated with microarrays for 60
min. A detection cocktail of 15 biotin-labeled phospho-specific
detection antibodies (R&D Systems) was used, with the concentration
of each antibody at 18 ng/mL. Colorimetric signals were detected by
transmission measurements with the ArraymateTM reader (Alere
Technology GmbH). Kinetic microarray data were analyzed with
rhodium(I)] (0.079 g, 0.14 mmol). Yield: 0.030 g (0.07 mmol, 25%),
1
3
yellow powder. H NMR (CDCl , ppm): 7.15 (d, 1H, J = 8.7 Hz,
3
4
3
4
ArH7), 6.83 (dd, 1 H, J = 2.3 Hz, J = 8.7 Hz, ArH6), 6.75 (d, 1H, J
2.3 Hz, ArH4), 5.14 (s, 2H, CH−COD), 4.26 (s, 6H, N−CH3),
.85 (s, 3H, −OCH ), 3.35 (s, 2H, CH−COD), 2.46 (m, 4H, CH −
=
3
3
2
13
1
COD), 2.01 (m, 4H, CH −COD). C NMR (CDCl ): 195.3 (d, J =
2
3
5
1
(
0.6 Hz, NHC), 156.4 (ArC-OCH ), 136.1 (ArC-3a), 130.0 (ArC-7a),
3
1
10.3 (ArC-7), 109.7 (ArC-6), 100.0 (d, J = 6.5 Hz, CH−COD), 94.0
3
0
1
1
KOMA software. Total protein concentrations were used for signal
normalization. The final results were calculated from two independent
experiments and are expressed as mean values with errors (standard
error of the mean, SEM).
ArC-4) 68.3 (d, J = 14.5 Hz, CH−COD), 68.2 (d, J = 14.5 Hz,
CH−COD), 56.0 (−OCH ), 35.0 (N−CH ), 35.0 (N−CH ), 32.9
3
3
3
(
(
(
CH −COD), 32.9 (CH −COD), 28.4 (CH −COD); MS(EI): 422
2
2
2
+
M ). Elemental analysis for C H ON ClRh (% calcd/found): C
18
24
2
Antiproliferative Effects in HT-29 and MDA-MB-231 Cells.
MCF-7 breast adenocarcinoma and HT-29 colon carcinoma cells were
maintained in DMEM High Glucose (PAA) supplemented with 50
mg/L gentamycin and 10% (v/v) fetal calf serum (FCS) prior to use.
The antiproliferative effects of the compounds were determined by the
crystal violet assay following an established procedure that was applied
51.14/51.54), H (5.72/5.63), N (6.63/6.56).
η ,η -Cycloocta-1,5-diene)iodido(1,3-dimethyl-5-methoxy-benzi-
midazol-2-ylidene)rhodium(I) (4d). 4b (0.079 g, 0.26 mmol), Ag O
2
2
(
2
2
2
(
0.047 g, 0.20 mmol), bis[iodido(η ,η -cycloocta-1,5-diene)rhodium-
(
I)] (0.095 g, 0.13 mmol). Yield: 0.032 g (0.06 mmol, 24%), light
1
3
brown powder. H NMR (CDCl , ppm): 7.13 (d, 1H, J = 8.2 Hz,
ArH7), 6.82 (dd, 1H, J = 3.0 Hz, J = 8.2 Hz, ArH6), 6.76 (d, 1H, J =
3
4
7−50
4
3
4
in a number of previous studies.
The results were calculated as
IC50 values obtained from two to three independent experiments and
are expressed as mean values ± SEM.
3
.0 Hz, ArH4), 5.32 (s, 2H, CH−COD), 4.16 (s, 6H, N−CH ), 3.85
3
(
2
(
s, 3H, −OCH ), 3.52 (s, 2H, CH−COD), 2.37 (m, 4H, CH −COD),
3
2
13
.04 (m, 2H, CH −COD), 1.86 (m, 2H, CH −COD). C NMR
2
2
1
CDCl ): 175.5 (d, J = 48.3 Hz, NHC), 156.3 (ArC-OCH ), 136.3
ASSOCIATED CONTENT
3
3
1
■
(
ArC-3a), 130.3 (ArC-7a), 110.1 (ArC-7), 109.6 (ArC-4), 97.9 (d, J =
1
*
S
Supporting Information
6
7
3
.3 Hz, CH−COD), 94.0 (ArC-6), 71.7 (d, J = 14.1 Hz, CH−COD),
1
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.5b01159.
1.7 (d, J = 14.1 Hz, CH−COD), 56.0 (−OCH ), 35.9 (N−CH ),
3
3
5.6 (N−CH ), 30.8 (CH −COD), 30.8 (CH −COD), 27.2 (CH −
3
2
2
2
+
COD); MS(EI): 514 (M ). Elemental analysis for C H ON IRh (%
18
24
2
calcd/found): C (42.02/42.25), H (4.70/4.79), N (5.45/4.95).
Mass Spectrometry. The compounds were dissolved at 10 mM in
dimethyl sulfoxide (DMSO, Sigma) and N,N-dimethylformamide
More details on mass spectrometry, effects on
mitochondria and cell metabolism, additional ELISA
array data (PDF)
(DMF, Acros), and the solutions were diluted with unbuffered water
MilliPore) to 100 μM and incubated in the dark at 37 °C. ESI mass
(
spectra in positive- and negative-ion modes were recorded after 10, 60,
20, 180, and 360 min and after 24 h. Aliquots were then further
AUTHOR INFORMATION
1
■
diluted to 5−10 μM prior to injection into the mass spectrometer.
Mass spectra were recorded on a Bruker AmaZon SL electrospray
ionization ion-trap (ESI-IT) mass spectrometer, and the resulting data
files were processed using Bruker Compass 1.3 and DataAnalysis 4.0.
The compounds were injected by direct infusion into the mass
spectrometer at 5 μL/min, and typical experimental conditions were as
follows: ±4.5 kV capillary voltage, 63% RF level, 55.2 trap drive, 180
Corresponding Author
E-mail: ingo.ott@tu-bs.de. Tel.: +49 531 3912743.
*
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
°C dry temperature, 8 psi nebulizer, 6 L/min dry gas. The average
The support by COST action CM1105 (Functional metal
complexes that bind to biomolecules) and the technical
assistance of Filip Groznica for performing some of the ESI-
MS measurements are gratefully acknowledged.
accumulation time in positive-ion mode was approximately 1 ms.
Experimental mass signals include a standard deviation of m/z ±0.02.
Cell Culture, Time-Course Experiments, and Lysate Prep-
aration for Microarray ELISA. Colon cancer cell line HT-29
(
ATCC) was maintained in DMEM High Glucose containing 10%
fetal calf serum (both from PAA Laboratories) at 37 °C with 5% CO .
2
ABBREVIATIONS USED
5
■
Cells were seeded in six-well plates (35 mm) at a density of 3 × 10
2
2
BCA, bicinchoninic acid; COD, η ,η -cycloocta-1,5-diene;
CREB, cAMP response element-binding protein; ERK,
extracellular-signal-regulated kinase; FCS, fetal calf serum;
cells/well and were grown for 24 h under standard cell culture
conditions to 60−70% confluence. For treatment, a stock solution of
2
c or 2d in dimethylformamide (DMF) was freshly prepared and
9
598
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
FAK, focal adhesion kinase; HSP, heat-shock protein; NHC, N-
heterocyclic carbene
rhodium(III) and iridium(III) complexes containing methyl-substi-
tuted phenanthroline ligands. ChemMedChem 2011, 6, 429−439.
(19) Geldmacher, Y.; Splith, K.; Kitanovic, I.; Alborzinia, H.; Can, S.;
Rubbiani, R.; Nazif, M. A.; Wefelmeier, P.; Prokop, A.; Ott, I.; Wolfl,
S.; Neundorf, I.; Sheldrick, W. S. Cellular impact and selectivity of half-
sandwich organorhodium(III) anticancer complexes and their
organoiridium(III) and trichloridorhodium(III) counterparts. JBIC, J.
Biol. Inorg. Chem. 2012, 17, 631−646.
REFERENCES
■
(
1) Hartinger, C. G.; Metzler-Nolte, N.; Dyson, P. J. Challenges and
opportunities in the development of organometallic anticancer drugs.
Organometallics 2012, 31, 5677−5685.
2) Gasser, G.; Ott, I.; Metzler-Nolte, N. Organometallic anticancer
compounds. J. Med. Chem. 2011, 54, 3−25.
3) Noffke, A. L.; Habtemariam, A.; Pizarro, A. M.; Sadler, P. J.
Designing organometallic compounds for catalysis and therapy. Chem.
Commun. 2012, 48, 5219−5246.
4) Oehninger, L.; Rubbiani, R.; Ott, I. N-Heterocyclic carbene metal
(
(20) Mollin, S.; Riedel, R.; Harms, K.; Meggers, E. Octahedral
rhodium(III) complexes as kinase inhibitors: Control of the relative
stereochemistry with acyclic tridentate ligands. J. Inorg. Biochem. 2015,
(
1
(
48, 11−21.
21) Ma, D. L.; Liu, L. J.; Leung, K. H.; Chen, Y. T.; Zhong, H. J.;
(
Chan, D. S.; Wang, H. M.; Leung, C. H. Antagonizing STAT3
dimerization with a rhodium(III) complex. Angew. Chem., Int. Ed.
complexes in medicinal chemistry. Dalton Trans. 2013, 42, 3269−
284.
5) Kascatan-Nebioglu, A.; Panzner, M. J.; Tessier, C. A.; Cannon, C.
3
(
2
(
014, 53, 9178−9182.
22) Liu, L. J.; Lin, S.; Chan, D. S.; Vong, C. T.; Hoi, P. M.; Wong, C.
L.; Youngs, W. J. N-Heterocyclic carbene−silver complexes: A new
Y.; Ma, D. L.; Leung, C. H. A rhodium(III) complex inhibits LPS-
induced nitric oxide production and angiogenic activity in cellulo. J.
Inorg. Biochem. 2014, 140, 23−28.
(23) Siu, F.-M.; Lin, I. W.-S.; Yan, K.; Lok, C.-N.; Low, K.-H.; Leung,
T. Y.-C.; Lam, T.-L.; Che, C.-M. Anticancer dirhodium(ii,ii)
carboxylates as potent inhibitors of ubiquitin-proteasome system.
Chem. Sci. 2012, 3, 1785−1793.
(24) Leung, C.-H.; Zhong, H.-J.; Chan, D. S.-H.; Ma, D.-L. Bioactive
iridium and hodium complexes as therapeutic agents. Coord. Chem.
Rev. 2013, 257, 1764−1776.
(25) Geldmacher, Y.; Oleszak, M.; Sheldrick, W. S. Rhodium(III) and
iridium(III) complexes as anticancer agents. Inorg. Chim. Acta 2012,
393, 84−102.
class of antibiotics. Coord. Chem. Rev. 2007, 251, 884−895.
(
6) Cisnetti, F.; Gautier, A. Metal/N-heterocyclic carbene complexes:
Opportunities for the development of anticancer metallodrugs. Angew.
Chem., Int. Ed. 2013, 52, 11976−11978.
̈
̌
̈
7) Cetinkaya, B.; Ozdemir, I.; Binbas i̧ oglu, B.; Durmaz, R.; Gunal, S.
(
Antibacterial and antifungal activities of complexes of ruthenium(II).
Arzneim. Forsch. 1999, 49, 538−540.
8) Cetinkaya, B.; Cetinkaya, E.; Ku
Antimicrobial activity of carbene complexes of rhodium(I) and
ruthenium(II). Arzneim. Forsch. 1996, 46, 821−823.
9) Durmaz, R.; Ku
(
̈
cu
̧
̈
kbay, H.; Durmaz, R.
(
̈
cu
̈
kbay, H.; Cetinkaya, E.; Cetinkaya, B.
Antimicrobial activity of rhodium (I) and ruthenium (II) carbene
complexes derived from benzimidazole against Staphylococcus aureus
isolates. Turk. J. Med. Sci. 1997, 27, 59−61.
(26) Kandioller, W.; Balsano, E.; Meier, S. M.; Jungwirth, U.; Goschl,
S.; Roller, A.; Jakupec, M. A.; Berger, W.; Keppler, B. K.; Hartinger, C.
G. Organometallic anticancer complexes of lapachol: Metal centre-
dependent formation of reactive oxygen species and correlation with
cytotoxicity. Chem. Commun. 2013, 49, 3348−3350.
(
10) Skander, M.; Retailleau, P.; Bourrie, B.; Schio, L.; Mailliet, P.;
Marinetti, A. N-Heterocyclic carbene-amine Pt(II) complexes, a new
chemical space for the development of platinum-based anticancer
drugs. J. Med. Chem. 2010, 53, 2146−2154.
(
11) Sun, R. W.-Y.; Chow, A. L.-F.; Li, X.-H.; Yan, J. J.; Chui, S. S.-Y.;
(27) Rubbiani, R.; Kitanovic, I.; Alborzinia, H.; Can, S.; Kitanovic, A.;
Che, C.-M. Luminescent cyclometalated platinum(II) complexes
containing N-heterocyclic carbene ligands with potent in vitro and
in vivo anti-cancer properties accumulate in cytoplasmic structures of
cancer cells. Chem. Sci. 2011, 2, 728−736.
12) Zhang, J.-J.; Che, C.-M.; Ott, I. Caffeine derived platinum(II) N-
heterocyclic carbene complexes with multiple anti-cancer activities. J.
Organomet. Chem. 2015, 782, 37−41.
13) Oehninger, L.; Stefanopoulou, M.; Alborzinia, H.; Schur, J.;
Ludewig, S.; Namikawa, K.; Mun
Onambele, L. A.; Stefanopoulou, M.; Geldmacher, Y.; Sheldrick, W. S.;
Wolber, G.; Prokop, A.; Wolfl, S.; Ott, I. Benzimidazol-2-ylidene
gold(I) complexes are thioredoxin reductase inhibitors with multiple
antitumor properties. J. Med. Chem. 2010, 53, 8608−8618.
(
(
28) Rubbiani, R.; Can, S.; Kitanovic, I.; Alborzinia, H.;
Stefanopoulou, M.; Kokoschka, M.; Monchgesang, S.; Sheldrick, W.
S.; Wolfl, S.; Ott, I. Comparative in vitro evaluation of N-heterocyclic
carbene gold(I) complexes of the benzimidazolylidene type. J. Med.
Chem. 2011, 54, 8646−8657.
(
̃
̈
oz-Castro, A.; Koster, R. W.;
Baumann, K.; Wolfl, S.; Sheldrick, W. S.; Ott, I. Evaluation of arene
̈
(
29) Matsuzawa, A.; Ichijo, H. Stress-responsive protein kinases in
redox-regulated apoptosis signaling. Antioxid. Redox Signaling 2005, 7,
72−481.
30) Charette, S. J.; Lavoie, J. N.; Lambert, H.; Landry, J. Inhibition
of Daxx-mediated apoptosis by heat shock protein 27. Mol. Cell. Biol.
000, 20, 7602−7612.
31) Mierke, C. T. The role of focal adhesion kinase in the regulation
of cellular mechanical properties. Phys. Biol. 2013, 10, 065005.
32) Berven, L. A.; Crouch, M. F. Cellular function of p70S6K: A role
in regulating cell motility. Immunol. Cell Biol. 2000, 78, 447−451.
33) Pi, X.; Yan, C.; Berk, B. C. Big mitogen-activated protein kinase
BMK1)/ERK5 protects endothelial cells from apoptosis. Circ. Res.
004, 94, 362−369.
34) Doble, B. W.; Woodgett, J. R. GSK-3: Tricks of the trade for a
multi-tasking kinase. J. Cell Sci. 2003, 116, 1175−1186.
35) Ott, I.; Schmidt, K.; Kircher, B.; Schumacher, P.; Wiglenda, T.;
ruthenium(II) N-heterocyclic carbene complexes as organometallics
interacting with thiol and selenol containing biomolecules. Dalton
Trans. 2013, 42, 1657−1666.
4
(
(
14) Gothe, Y.; Marzo, T.; Messori, L.; Metzler-Nolte, N. Cytotoxic
activity and protein binding through an unusual oxidative mechanism
by an iridium(I)-NHC complex. Chem. Commun. 2015, 51, 3151−
153.
2
(
3
(
15) Simpson, P. V.; Schmidt, C.; Ott, I.; Bruhn, H.; Schatzschneider,
(
U. Synthesis, cellular uptake and biological activity against pathogenic
microorganisms and cancer cells of rhodium and iridium N-
heterocyclic carbene complexes bearing charged substituents. Eur. J.
Inorg. Chem. 2013, 2013, 5547−5554.
(
(
2
(
16) McConnell, J. R.; Rananaware, D. P.; Ramsey, D. M.; Buys, K.
(
N.; Cole, M. L.; McAlpine, S. R. A potential rhodium cancer therapy:
Studies of a cytotoxic organorhodium(I) complex that binds DNA.
Bioorg. Med. Chem. Lett. 2013, 23, 2527−2531.
(
Gust, R. Antitumor-active cobalt-alkyne complexes derived from
acetylsalicylic acid: Studies on the mode of drug action. J. Med. Chem.
2005, 48, 622−629.
(36) Schur, J.; Manna, C. M.; Deally, A.; Koster, R. W.; Tacke, M.;
Tshuva, E. Y.; Ott, I. A comparative chemical-biological evaluation of
titanium(IV) complexes with a salan or cyclopentadienyl ligand. Chem.
Commun. (Cambridge, U. K.) 2013, 49, 4785−4787.
(
17) Oehninger, L.; Kuster, L. N.; Schmidt, C.; Munoz-Castro, A.;
Prokop, A.; Ott, I. A chemical-biological evaluation of rhodium(I) N-
heterocyclic carbene complexes as prospective anticancer drugs.
Chem.Eur. J. 2013, 19, 17871−17880.
(
18) Geldmacher, Y.; Kitanovic, I.; Alborzinia, H.; Bergerhoff, K.;
Rubbiani, R.; Wefelmeier, P.; Prokop, A.; Gust, R.; Ott, I.; Wolfl, S.;
Sheldrick, W. S. Cellular selectivity and biological impact of cytotoxic
9
599
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600

Journal of Medicinal Chemistry
Article
(
37) Wang, D.; Lippard, S. J. Cellular processing of platinum
anticancer drugs. Nat. Rev. Drug Discovery 2005, 4, 307−320.
38) Holenya, P.; Heigwer, F.; Wolfl, S. KOMA: ELISA-microarray
(
calibration and data analysis based on kinetic signal amplification. J.
Immunol. Methods 2012, 380, 10−15.
(
39) Holenya, P.; Kitanovic, I.; Heigwer, F.; Wolfl, S. Microarray-
based kinetic colorimetric detection for quantitative multiplex protein
phosphorylation analysis. Proteomics 2011, 11, 2129−2133.
(
40) Alborzinia, H.; Can, S.; Holenya, P.; Scholl, C.; Lederer, E.;
Kitanovic, I.; Wolfl, S. Real-time monitoring of cisplatin-induced cell
death. PLoS One 2011, 6, e19714.
(
41) Holenya, P.; Can, S.; Rubbiani, R.; Alborzinia, H.; Junger, A.;
Cheng, X.; Ott, I.; Wolfl, S. Detailed analysis of pro-apoptotic signaling
and metabolic adaptation triggered by a N-heterocyclic carbene-
gold(I) complex. Metallomics 2014, 6, 1591−1601.
(
42) Cheng, X.; Holenya, P.; Can, S.; Alborzinia, H.; Rubbiani, R.;
Ott, I.; Wolfl, S. A TrxR inhibiting gold(I) NHC complex induces
apoptosis through ASK1-p38-MAPK signaling in pancreatic cancer
cells. Mol. Cancer 2014, 13, 221.
(
43) Romero-Canelon, I.; Salassa, L.; Sadler, P. J. The contrasting
activity of iodido versus chlorido ruthenium and osmium arene azo-
and imino-pyridine anticancer complexes: Control of cell selectivity,
cross-resistance, p53 dependence, and apoptosis pathway. J. Med.
Chem. 2013, 56, 1291−1300.
(
44) Karuvalam, R. P.; Haridas, K. R.; Shetty, S. N. Trimethylsilyl
chloride catalyzed synthesis of substituted benzimidazoles using two
phase system under microwave conditions, and their antimicrobial
studies. J. Chil. Chem. Soc. 2012, 57, 1122−1125.
(
45) Eren, B.; Erdogan, G. Eco-friendly and efficient synthesis of
benzimidazole derivatives using iron oxide modified sepiolite catalyst.
React. Kinet., Mech. Catal. 2012, 107, 333−344.
(
46) Reddy, K. K.; Subba Rao, N. V. Alkylation and aralkylation of N-
heterocycles. Part II. Methylation and benzylation of 5 (or 6)-methyl
benzimidazoles. Proc. Ind. Acad. Sci, A 1969, 70, 81−88.
(
47) Rubbiani, R.; Salassa, L.; de Almeida, A.; Casini, A.; Ott, I.
Cytotoxic gold(I) N-heterocyclic carbene complexes with phosphane
ligands as potent enzyme inhibitors. ChemMedChem 2014, 9, 1205−
1210.
(
48) Navakoski de Oliveira, K.; Andermark, V.; Onambele, L. A.;
Dahl, G.; Prokop, A.; Ott, I. Organotin complexes containing
carboxylate ligands with maleimide and naphthalimide derived partial
structures: TrxR inhibition, cytotoxicity and activity in resistant cancer
cells. Eur. J. Med. Chem. 2014, 87, 794−800.
(
49) Serebryanskaya, T. V.; Lyakhov, A. S.; Ivashkevich, L. S.; Schur,
J.; Frias, C.; Prokop, A.; Ott, I. Gold(I) thiotetrazolates as thioredoxin
reductase inhibitors and antiproliferative agents. Dalton Trans 2015,
4
(
4, 1161−1169.
50) Serebryanskaya, T. V.; Zolotarev, A. A.; Ott, I. A novel
aminotriazole-based NHC complex for the design of goldIJI) anti-
cancer agents: Synthesis and biological evaluation. MedChemComm
2
015, 6, 1186−1189.
9
600
DOI: 10.1021/acs.jmedchem.5b01159
J. Med. Chem. 2015, 58, 9591−9600