Cell-based studies of the first-in-class half-sandwich Ir(III) complex containing histone deacetylase inhibitor 4-phenylbutyrate

Article abstract of DOI:10.1002/aoc.4246

We report on a cytotoxic half-sandwich iridium(III) complex [Ir(η⁵-Cp^ph)(phen)(PB)]PF₆ (1-PB), containing a monodentate coordinated O-donor 4-phenylbutyrato ligand (PB) belonging to the family of histone deacetylase inhibitors (HDACi); HCp^ph = (2,3,4,5-tetramethylcyclopenta-2,4-dien-1-yl)benzene, phen = 1,?10-phenanthroline. The solution behaviour studies indicated that complex 1-PB partially hydrolysed in the mixture of methanol and water (1:4, v/v), resulting in the release of the PB ligand. The extent of the PB ligand release increased in the presence of 2 molar equiv. of the reduced glutathione (GSH). Complex 1-PB exhibited comparable in vitro cytotoxicity against the cisplatin-sensitive (IC₅₀ = 15.8?μM) and -resistant (IC₅₀ = 13.0?μM) variants of the A2780 human ovarian carcinoma cells, while its potency against the MRC-5 human normal fibroblast cells was markedly lower (IC₅₀ = 124.1?μM). The cytotoxicity studies revealed an ability of complex 1-PB to overcome the acquired resistance against cisplatin, with the resistance factor (RF = 0.8) being markedly lower than for complex 1-Cl (RF = 1.8) and cisplatin (RF = 2.9). The A2780 cell-based flow cytometry experiments showed different cell cycle modification induced by complex 1-PB and cisplatin, induction of production of reactive oxygen species, and higher mitochondria membrane potential depleted cell populations after the treatment by complex 1-PB as compared with cisplatin. In the cell-free assay, complex 1-PB inhibited the HDAC activity to ca 66% as compared to ca 74% valid for NaPB. The [Ir(η⁵-Cp^ph)(phen)(H₂O)]²⁺ species (1-OH₂), representing the hydrolysis product of both complexes 1-PB and 1-Cl, induced hydroxyl radical from the hydrogen peroxide, as proved by the EPR spin trapping studies with the 5-(diethoxyphosphoryl)-5-methyl-1-pyrroline-N-oxide (DEPMPO) spin trap.

Full text of DOI:10.1002/aoc.4246

Received: 6 September 2017
DOI: 10.1002/aoc.4246
Revised: 4 December 2017
Accepted: 4 December 2017
F U L L P A P E R
Cell‐based studies of the first‐in‐class half‐sandwich Ir(III)
complex containing histone deacetylase inhibitor
4‐phenylbutyrate
Pavel Štarha1
| Zdeněk Trávníček1
| Bohuslav Drahoš1 | Radovan Herchel1 |
Zdeněk Dvořák2
1 Department of Inorganic Chemistry &
Regional Centre of Advanced
We report on a cytotoxic half‐sandwich iridium(III) complex [Ir(η5‐Cpph)(phen)
Technologies and Materials, Faculty of
Science, Palacký University in Olomouc,
17. listopadu 12, 771 46 Olomouc, Czech
Republic
2 Department of Cell Biology and Genetics
& Regional Centre of Advanced
Technologies and Materials, Faculty of
Science, Palacký University in Olomouc,
Šlechtitelů 27, 783 71 Olomouc, Czech
Republic
(PB)]PF6 (1‐PB), containing
a
monodentate coordinated O‐donor 4‐
phenylbutyrato ligand (PB) belonging to the family of histone deacetylase
inhibitors (HDACi); HCpph = (2,3,4,5‐tetramethylcyclopenta‐2,4‐dien‐1‐yl)ben-
zene, phen = 1,10‐phenanthroline. The solution behaviour studies indicated that
complex 1‐PB partially hydrolysed in the mixture of methanol and water
(1:4, v/v), resulting in the release of the PB ligand. The extent of the PB ligand
release increased in the presence of 2 molar equiv. of the reduced glutathione
(GSH). Complex 1‐PB exhibited comparable in vitro cytotoxicity against the
cisplatin‐sensitive (IC50 = 15.8 μM) and ‐resistant (IC50 = 13.0 μM) variants of the
A2780 human ovarian carcinoma cells, while its potency against the MRC‐5 human
normal fibroblast cells was markedly lower (IC50 = 124.1 μM). The cytotoxicity
studies revealed an ability of complex 1‐PB to overcome the acquired resistance
against cisplatin, with the resistance factor (RF = 0.8) being markedly lower than
for complex 1‐Cl (RF = 1.8) and cisplatin (RF = 2.9). The A2780 cell‐based flow
cytometry experiments showed different cell cycle modification induced by
complex 1‐PB and cisplatin, induction of production of reactive oxygen species,
and higher mitochondria membrane potential depleted cell populations after the
treatment by complex 1‐PB as compared with cisplatin. In the cell‐free assay,
complex 1‐PB inhibited the HDAC activity to ca 66% as compared to ca 74% valid
for NaPB. The [Ir(η5‐Cpph)(phen)(H2O)]2+ species (1‐OH2), representing the
hydrolysis product of both complexes 1‐PB and 1‐Cl, induced hydroxyl radical
from the hydrogen peroxide, as proved by the EPR spin trapping studies with
the 5‐(diethoxyphosphoryl)‐5‐methyl‐1‐pyrroline‐N‐oxide (DEPMPO) spin trap.
Correspondence
Zdeněk Trávníček, Department of
Inorganic Chemistry, Faculty of Science,
Palacký University in Olomouc, 17.
listopadu 12, 771 46 Olomouc, Czech
Republic.
Email: zdenek.travnicek@upol.cz
Funding information
The Ministry of Education, Youth and
Sports of the Czech Republic, Grant/
Award Number: CZ.1.05/2.1.00/19.0377
LO1305; The Czech Science Fundation,
Grant/Award Number: 17‐08512Y
KEYWORDS
4‐phenylbutyrate, cytotoxicity, flow cytometry, half‐sandwich, iridium(III) complexes
1 | INTRODUCTION
resistance[1] incessantly motivate bioinorganic chemists
to develop novel types of compounds for more efficient
The negative side effects of platinum‐based anticancer
chemotherapy as well as the problems caused by
and safer therapy. Among them, agents consisting of
potent transition metal species containing a releasable
Appl Organometal Chem. 2018;e4246.
wileyonlinelibrary.com/journal/aoc
Copyright © 2018 John Wiley & Sons, Ltd.
1 of 11
https://doi.org/10.1002/aoc.4246

2 of 11
ŠTARHA ET AL.
bioactive ligand acting through a different mechanism of
action, seem to be a perspective strategy for the medicinal
chemistry,[2] as recently reported for several Pt(IV) com-
plexes, containing for example (D)‐1‐methyltryptophan,
4‐phenylbutyrate or valproate as the bioactive axial
ligands.[3]
cytometry studies of some relevant processes connected
with its mechanism of action. To the best of our knowl-
edge, complex 1‐PB represents the first half‐sandwich
iridium(III) complex containing a monodentate O‐donor
ligand bearing its own biological activity.
Many types of non‐platinum complexes,[4] including
half‐sandwich cyclopentadienyl (Cpx) Ir(III) complexes
[Ir(η5‐Cpx)(X^Y)Cl]+/0, have been reported as anticancer
active against various tumours; X^Y = a bidentate‐coordi-
nated ligand, e.g., N,N‐donor 1,10‐phenanthroline (phen)
or C,N‐donor deprotonated 2‐phenylpyridine.[5] The
mechanism of action of these Ir(III) agents is still not fully
understood but relevant evidences have been reported for
a dual‐targeting action combining the nuclear DNA
damage and a redox‐mediated effect connected with the
mitochondria dysfunction and induction of the reactive
oxygen species (ROS).[5c,6] Only a few works have studied
the effect of the replacement of the chlorido ligand in the
half‐sandwich Ir(III)‐chlorido complexes by a different
one (pyridine and its derivatives[7] or hydrosulfide[8]) on
the anticancer activity.
In this work, the chlorido ligand of the formerly
reported anticancer active complex [Ir(η5‐Cpph)(phen)Cl]
PF6 (1‐Cl)[5a] was replaced by 4‐phenylbutyrate (PB),
resulting in complex [Ir(η5‐Cpph)(phen)(PB)]PF6 (1‐PB;
Figure 1). Sodium 4‐phenylbutyrate (NaPB) is an FDA
approved drug for the treatment of patients with urea
cycle disorders (hyperammonemia) and a clinically inves-
tigated anticancer drug for the treatment of various types
of tumours, which belongs to the family of histone
deacetylase inhibitors (HDACi).[9] We report on the prep-
aration, thorough characterization and studies of solution
behaviour and in vitro cytotoxicity of complex 1‐PB
against both cancer and healthy cells, and cell‐based flow
2 | EXPERIMENTAL SECTION
2.1 | Materials
IrCl3·nH2O (99%) was purchased from Precious Metals
Online. Other chemicals, in particular 1,10‐
phenanthroline (≥99.0%), 4‐phenylbutyric acid (PBA;
99.0%), sodium hydroxide (≥98.0%), silver trifluoro-
methanesulfonate (AgOTf; ≥98.0%), 2,3,4,5‐tetramethyl‐
2‐cyclopentenone (95%), phenylmagnesium bromide
(3.0 M in diethyl ether), MgSO4 (≥99.5%), NH4PF6
(≥98%), reduced glutathione (GSH; ≥98.0%), (SP‐4‐2)‐
diammine‐dichloridoplatinum(II) (cisplatin; ≥99.9%),
phosphate buffered saline (PBS), hydrogen peroxide
(30.0 wt. % in H2O) and ammonium formate (AF; for
HPLC, ≥99.0%) were purchased from VWR International
(Stříbrná Skalice, Czech Republic) and Sigma‐Aldrich
(Prague, Czech Republic). Solvents of laboratory (for syn-
theses and characterization) and HPLC (H2O and MeCN)
grades, and deuterated solvents for NMR (DMSO‐d6,
MeOD‐d4, D2O) were supplied by VWR International,
Sigma‐Aldrich, Fisher Scientific (Pardubice, Czech
Republic) and Litolab (Chudobín, Czech Republic). Sol-
vents were used without further purification, except for
THF which was dried by overnight standing over 4 Å
molecular sieves. Roswell Park Memorial Institute
(RPMI‐1640) medium, trypsin and phosphate‐buffered
saline (PBS) were purchased from Sigma‐Aldrich and
Fisher Scientific, while 5‐(diethoxyphosphoryl)‐5‐methyl‐
FIGURE 1 Structural formula of
1
complex 1‐PB and its H NMR spectrum
given with the assignment of the observed
signals

ŠTARHA ET AL.
3 of 11
1‐pyrroline‐N‐oxide (DEPMPO) was purchased from
Enzo Life Sciences (Farmingdale, NY, USA). Complexes
doublets and m = multiplet. Mass spectra were obtained
on the methanol solutions using the positive ion mode
of electrospray ionization technique (ESI+) by LCQ Fleet
ion trap spectrometer (Thermo Fisher Scientific,
Waltham, MA, USA) and the data were analysed by
QualBrowser software (version 2.0.7). Infrared spectros-
copy (400–4000 cm–1, Attenuated total reflectance (ATR)
technique) was performed by a Nexus 670 FT‐IR
spectrometer (Thermo Fisher Scientific). Elemental
analysis determining a content of C, H and N was
performed by a Flash 2000 CHNS Elemental Analyzer
(Thermo Fisher Scientific). Reversed‐phase high‐perfor-
mance liquid chromatography (RP‐HPLC) coupled to
the positive electrospray ionization mode mass spectrom-
etry (ESI+ MS) was carried out using UHPLC‐MS
(Dionex/Thermo Fisher Scientific) equipped with an
ReproSil‐Pur Basic C18 (5 μm pore size, 200 × 4.6 mm).
The mixture of 0.1% AF in H2O (A) and acetonitrile
(MeCN; B) was used as the mobile phase at gradients of
10 % B (t = 0 min), 80 % B (t = 30 min), 80 % B (t =
40 min), 10 % B (t = 41 min) and 10 % B (t = 55 min) over
a 55 min period (1 ml min−1 flow rate). The detection
wavelength was 254 nm.
[Ir(μ‐Cl)(η5‐Cpph)Cl]2
and [Ir(η5‐Cpph)(phen)Cl]PF6
[10]
(1‐Cl)5a were prepared using the modification of the
reported procedures. The synthesis of silver(I) 4‐
phenylbutyrate (AgPB) is described elsewhere.[11]
2.2 | Synthesis of complex 1‐PB
The formerly reported[5a] complex 1‐Cl (0.05 mmol; see
supporting information) reacted overnight in methanol
(10 mL) with an excess (0.15 mmol) of AgPB (ambient
temperature, protected from light by aluminium foil).
After that, a white solid (AgCl) was filtered off and
NH4PF6 (0.25 mmol) was added to the obtained solution.
The solvent volume was reduced to ca 0.5 mL and diethyl
ether (ca 5 ml) was poured in. The yellow precipitate was
collected, washed with diethyl ether (3 × 1 mL), and dried
under vacuum in desiccator. Yield: 65% (related to com-
1
plex 1‐Cl). H NMR (600 MHz, DMSO‐d6): δ 9.23 (d, 2H,
JHH = 4.6 Hz, phen), 8.95 (d, 2H, JHH = 8.3 Hz, phen),
8.30 (s, 2H, phen), 8.15 (dd, 2H, JHH = 7.8, 5.0 Hz, phen),
7.53–7.41 (m, 5H, Cppharom.), 7.04 (m, 3H, PBarom.), 6.46
(m, 2H, PBarom.), 1.88 (s, 6H, Cpphaliph.), 1.70 (s, 6H,
Cpphaliph.), 1.64 (m, 4H, PBaliph.), 1.10 (m, 2H, PBaliph.
)
2.4 | Solution chemistry and interaction
with GSH
ppm. 13C NMR (150 MHz, DMSO‐d6): δ 177.6, 152.9,
147.1, 141.8, 140.1, 131.0–129.5, 128.5, 128.4, 127.5,
126.0, 95.1, 87.2, 35.3, 34.1, 28.1, 9.9, 9.0 ppm. Anal. Calcd.
for IrC37H36N2O2PF6: C, 50.62; H, 4.13; N, 3.19%; found:
C, 50.29; H, 4.09; N, 2.94%.
The appropriate amount of complex 1‐PB (and 1‐Cl for
comparative purposes) for 1 mM solutions was dissolved
in 120 μL of MeOD‐d4 and either 480 μL of D2O (for the
studies of hydrolysis) or a GSH (2 molar equiv.) solution
in 480 μL of D2O (for GSH interactions) was added. Ana-
logical experiments were performed in the mixtures of
120 μL of MeOD‐d4 and 480 μL of D2O solution of PBS
(pH = 7.2). Further, the GSH interaction study of complex
1‐PB was carried out in the mixture of 120 μL of MeOD‐d4
and 480 μL of D2O, whose pH value was adjusted by 0.1%
The [Ir(η5‐Cpph)(phen)(H2O)]2+ (1‐OH2) and [Ir(η5‐
Cpph)(phen)(D2O)]2+ (1‐OD2) species were prepared from
complex 1‐Cl by its reaction with the stoichiometric
amount of AgOTf in either DMF (for 1‐OH2) or MeOD‐
d4 (for 1‐OD2). After 30 min of shaking (Vibramax 100,
Heidolph Instruments), the formed AgCl was removed
by centrifugation (10,000 rpm for 3 min), and the obtained
solutions of the 1‐OH2 and 1‐OD2 species were stored in
the fridge until required.
1
KOD to 7.2. H NMR spectra were recorded at different
time points over 48 h (300 K, JEOL JNM‐ECA 600II).
1
The obtained H NMR spectra were calibrated against
the residual signal of D2O found at 4.85 ppm.[12] For
comparative purposes, the 1H NMR spectra were also
recorded for 1‐OD2, free GSH and NaPB dissolved in the
same mixture of solvents (20% MeOD‐d4/80% D2O with-
out or with PBS). Similar experiments were carried out
in the mixture of 20% MeOH/80% H2O and analysed by
ESI+ and RP‐HPLC/ESI+ mass spectrometries.
Complex 1‐OD2 (3 × 120 μL in MeOD‐d4) was diluted
by D2O to give 3 × 600 μL of 1 mM solution in 20% MeOD‐
d4/80% D2O. pH of two solutions was adjusted by an
addition of 0.1% DClO4 and 0.1% KOD to the values of 4
and 10, respectively, and the 1H NMR spectra were
recorded (300 K, JEOL JNM‐ECA 600II).
2.3 | Methods
1
1
1
1H, 13C, H‐1H gsCOSY, APT, H‐13C gsHMQC and H‐
13C gsHMBC spectra were recorded by a JNM‐ECA
600II spectrometer (JEOL Ltd., Tokyo, Japan) at 300 K;
gs = gradient selected, COSY = correlation spectroscopy,
APT = Attached Proton Test, HMQC = heteronuclear
multiple quantum coherence, HMBC = heteronuclear
multiple bond coherence. H and 13C NMR spectra were
1
calibrated against the residual signals of DMSO‐d6, found
at 2.50 ppm (1H) and 39.52 ppm (13C).[12] The splitting of
the proton resonances in the reported 1H spectra is
defined as s = singlet, d = doublet, dd = doublet of

4 of 11
ŠTARHA ET AL.
centrifugation and the obtained pellets were immediately
digested in 500 μL of nitric acid (3 min, 150 °C) using a
microwave reaction system Monowave 300 (Anton Paar
GmbH, Graz, Austria). The solutions were diluted with
4.5 mL of water and the Ir content was determined by
inductively coupled plasma mass spectrometry (ICP‐MS)
using the 7700x ICP‐MS device (Agilent Technologies).
The obtained values were corrected for adsorption effects.
The experiments were conducted in triplicate and the
results are presented as arithmetic mean SD.
2.5 | Cell culture
The human ovarian carcinoma cell line (A2780), cisplatin‐
resistant human ovarian carcinoma cell line (A2780R)
and human normal lung fibroblast cell line (MRC‐5) were
cultured according to the supplier's (European Collection
of Cell Cultures, ECACC) instructions, using RPMI‐1640
medium supplemented with 10% of fetal calf serum, 1%
of 2 mM glutamine and 1% penicillin/streptomycin. All
cell lines were grown as adherent monolayers at 37 °C
and 5% CO2 in a humidified atmosphere. The 200 mM
stock solution of complex 1‐PB was prepared by its
dissolving in 500 μL of DMF. NaPB and cisplatin were
involved in the testing for comparative purposes.
2.5.3 | Cell cycle analysis
The A2780 cells were pre‐incubated in the 6‐well plates
(1.0 × 106 cells per well) for 24 h at 37 °C and 5% CO2
atmosphere. Complex 1‐PB (and cisplatin for comparative
purposes) were added at the equipotent concentrations
(equal to their IC50). After 24 h exposure time, the cells
were harvested using trypsin, washed twice with PBS
and fixed in 70% ethanol. The resuspended cells were
stained for 30 min in the dark using a solution of
propidium iodide (PI) supplemented with RNase A
(25 °C). After that, the cells were washed by PBS, resus-
pended to PBS and the DNA content was assessed using
flow cytometry (CytoFLEX Flow Cytometer, Beckman
Coulter) detecting emission of DNA‐bound PI (maximum
at 617 nm) after excitation at 535 nm. Untreated A2780
cells were used as a negative control. The experiments
were conducted in triplicate and the data were analysed
using CytExpert Software.
2.5.1 | In vitro cytotoxicity
The cultured A2780, A2780R and MRC‐5 cells were
seeded to 96‐well culture plates, pre‐incubated in drug‐
free media at 37 °C for 24 h and then treated for next
24 h at 37 °C with different concentrations of complexes
1‐PB and 1‐Cl, NaPB and cisplatin prepared from their
stock solutions by dilution with RPMI‐1640 medium.
After 24 h drug exposure, the supernatants were removed
and the cells were washed with drug‐free PBS followed by
72 h recovery in drug‐free medium at 37 °C. An MTT
assay (MTT = 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyl-
tetrazolium bromide) was used to determine the cell
viability. A concentration of the formazan formed from
MTT by the mitochondrial reduction of the viable cells,
was evaluated spectrophotometrically at 540 nm using
an Infinite 200 PRO microplate reader (Tecan Group
Ltd., Männedorf, Switzerland).
The data were expressed as the percentage of viability,
where 100% and 0% represents the treatments with
negative (0.1% DMF in medium) and positive (1% Triton
X‐100) controls, respectively. The data from the cancer
cells were acquired from three independent experiments
(conducted in triplicate) using cells from different
passages. The resulting IC50 values (μM) were calculated
from the viability curves and the results are presented as
arithmetic mean standard deviation (SD).
2.5.4 | Mitochondrial membrane potential
assay
The A2780 cells were treated by the IC50 concentrations of
complex 1‐PB for 24 h (37 °C, 5% CO2) and processed as
described in the supplied MITO‐ID® Membrane potential
detection kit (emission maxima at 590 and 525 nm for
the orange and green dyes, respectively) from Enzo Life
Sciences. The negative control included the untreated
A2780 cells, while the cells treated with carbonyl cyanide
3‐chlorophenylhydrazone (CCCP; 2 μM final concentra-
tion) were used as the positive control. The cells were
analysed by flow cytometry on a CytoFlex device
(Beckman Coulter). The experiments were conducted in
triplicate (ca. 3 × 105 cells per samples).
2.5.2 | Cellular accumulation
The cultured A2780 cells were seeded in 6‐well culture
plates (1 × 106 cells per well) and incubated overnight
(37 °C and 5% CO2 in a humidified incubator). After that
the cells were treated with the IC50 concentrations of
complexes 1‐PB and 1‐Cl. The treatment was followed
by washing with PBS (2 × 2 mL) and the washed cells
were harvested by the trypsin treatment, collected and
centrifuged in PBS. The supernatants were discarded after
2.5.5 | ROS/Superoxide assay
The pre‐incubated A2780 cells (ca. 3.0 × 105 per sample)
were exposed to the IC50 concentrations of complexes 1‐
PB and 1‐Cl for 2 or 24 h (37 °C, 5% CO2). After that,
the solutions with complexes 1‐PB and 1‐Cl were

ŠTARHA ET AL.
5 of 11
removed from the cells, which were subsequently stained
according the instruction of the used Total ROS/Superox-
ide detection kit (Enzo Life Sciences). The cells treated
with pyocyanin were used as the positive control, while
the negative control is represented by untreated cells.
Flow cytometry experiments were performed on a
CytoFlex device (Beckman Coulter) detecting emissions
at 525 nm (total oxidative stress) and 620 nm (superox-
ide). The experiments were conducted in triplicate.
3 | RESULTS AND DISCUSSION
3.1 | Syntheses and chemistry
The design of complex 1‐PB was based on the
replacement of the chlorido ligand of the cytotoxic half‐
sandwich Ir(III)‐chlorido complex, [Ir(η5‐Cpph)(phen)Cl]
PF6 (1‐Cl),[5a] by a bioactive ligand (4‐phenylbutyrate).
On the other way around, complex 1‐PB was assumed
to behave as a prodrug releasing two independent cyto-
toxic species under the (pseudo)physiological conditions.
This is important to accentuate as the crucial difference
from previously reported half‐sandwich Ir(III) complexes
containing the bioactive ligand coordinated on the
inactive Ir(III) species. This can be exemplified on [Ir(η5‐
2.6 | Histone deacetylase inhibition in
cell‐free media
The effect of complex 1‐PB and reference compounds
(NaPB and cisplatin) on the activity of histone deacetylase
(HDAC) was evaluated using a fluorometric Histone
Deacetylase Assay Kit (Sigma‐Aldrich, Prague, Czech
Republic). The experiments were done according the
supplier's instructions. The samples were incubated with
the IC50 concentrations of complex 1‐PB and cisplatin
(from the cytotoxicity experiments) as well as with
16 μM NaPB at 30 °C for 30 min. Fluorescence (excitation
at 360 nm) was recorded at 460 nm by an Infinite 200 PRO
microplate reader.
Cp*)(cur)(PTA)]+ containing curcumin (cur) as
cytotoxic ligand, while the metal‐based hydrolysates
[Ir(η5‐Cp*)Cl(OD2)(PTA)]+ and [Ir(η5‐Cp*)Cl2(PTA)]
were reported as inactive.[13]
a
Complex 1‐PB was prepared by the straightforward
reaction of the previously reported chlorido complex 1‐
Cl5a with an excess of silver(I) 4‐phenylbutyrate (AgPB),
as recently reported for Ru(II) complex [Ru(η6‐pcym)
(dpa)(PB)]PF6 (pcym = p‐cymene and dpa = 2,2`‐
dipyridylamine).[11] Complex 1‐PB was isolated in a good
yield and its 99.4% purity (tR = 9.97 min) was determined
by RP‐HPLC/ESI+ MS (supporting information, Figure
S1). The RP‐HPLC results also suggested higher relative
lipophilicity[14] for complex 1‐PB than for complex 1‐Cl
(tR = 8.16 min). The ESI+ mass spectrum of complex 1‐
PB contained peaks of the [Ir(Cpph)(phen)(PB)]+
(733.2 m/z, 100% relative intensity) and {[Ir(Cpph)
(phen)]–H}+ (569.2 m/z, 75% relative intensity) species
(supporting information, Figure 2). The FT‐IR spectrum
of complex 1‐PB showed the peaks assignable to the PB,
2.6.1 | EPR spin‐trapping (EPR‐ST) studies
The stock solutions of the studied compounds (1‐PB, 1‐Cl
or 1‐OH2; 10 mM solution in DMF), H2O2 (100 mM in
75 mM PBS) and DEPMPO (60 mM in 75 mM PBS) were
prepared, and 50 μL aliquots were mixed to give 150 μL of
samples. The mixtures of the studied compounds and
DEPMPO, or H2O2 and DEPMPO were also measured
for comparative purposes (blank controls). The measure-
ments were performed using a MiniScope MS5000
Magnettech X‐band EPR spectrometer (Freiberg Instru-
ments GmbH, Freiberg, Germany) at ambient tempera-
ture on fresh samples. All the samples were measured in
50 μL capillaries using the following settings: modulation
amplitude 0.1 mT, modulation frequency 100 kHz, micro-
wave power 20 mW, sweep width 25 mT, scan time 4 min.
The measured samples consisted of complexes 1‐PB,
1‐Cl or 1‐OH2 (final concentration of 3.3 mM), mixed
with H2O2 (final concentration of 33.0 mM) and DEPMPO
(final concentration of 16.7 mM) in the mixture of DMF
and 75 mM PBS (1:2, v/v).
phen and Cpph ligands, as well as to the PF6 counterion
–
(supporting information, Figure S2).[15] The 1H NMR phen
and Cpph signals of complex 1‐PB showed in the same
region as reported for complex 1‐Cl (supporting
1
information, Figure S3).[5a] The H NMR signals of the
aliphatic α‐CH2, β‐CH2 and γ‐CH2 hydrogens of free PB
anion (measured on NaPB) shifted upfield by 0.89 ppm,
0.60 ppm and 0.24 ppm, respectively, after its coordination
into the structure of complex 1‐PB, which is consistent
with the [Ru(η6‐pcym)(dpa)(PB)]PF6 complex.[11] Only a
small Δδ = 1.0 ppm was detected in the 13C NMR spectra
for the carboxyl carbon atom of the PB anion upon its
coordination to the Ir(III) atom of complex 1‐PB.
2.7 | Statistical analysis
An ANOVA test was used for statistical analysis with the
values of p < 0.05 (*), 0.01 (**) and 0.005 (***) considered
to be statistically significant. QC Expert 3.2 was used to
perform the analysis (TriloByte Statistical Software).
3.2 | Solution chemistry
1
Complex 1‐PB readily hydrolysed, as proved by H NMR
(Figure 3) and RP‐HPLC/ESI+ MS (supporting

6 of 11
ŠTARHA ET AL.
FIGURE 2 ESI+ mass spectrum of
complex 1‐PB given with the comparison
of the experimental (top) and simulated
(bottom) isotopic distribution of the
{[Ir(Cpph)(phen)]–H}+ (blue) and [Ir(Cpph
(phen)(PB)]+ (green) species
)
FIGURE 3 1H NMR spectra (20%
MeOD‐d4/80% D2O; time = 0, 0.5, 1, 4 and
48 h) showing the hydrolysis of 1‐PB (top),
and a progress of the PB ligand release (1H
NMR studies), as observed at different
time points for the experiments performed
at different pH values and without or with
2 mol equiv. of GSH (bottom). Presence (+)
or absence (–) of phosphate buffered saline
(PBS), reduced glutathione (GSH) and/or
0.1% KOD in D2O in the studied mixtures
is symbolized by the + and – symbols. The
pH values equal 7.0 for –/–/–, 7.2 for +/–/
–, 4.7 for –/+/–, and 7.2 for +/+/– and –/
+/+ (given as PBS/GSH/KOD)
1
information, Figure S4) experiments. Concerning the H
NMR results, a gradual release of PB and [Ir(η5‐Cpph)
(phen)(OD2)]2+ (1‐OD2; supporting information, Figure
S4 and S5) was observed for complex 1‐PB in 20%
MeOD‐d4/80% D2O (pH = 7.0), which reached the equilib-
rium state after ca 4 h of standing at ambient temperature,
when the rate of the PB ligand release reached approxi-
mately 60%. In the PBS solution (pH = 7.2), the extent of
the PB ligand release was higher (85% after 4 h) and
reached even 100% after 24 h of standing at ambient
temperature (Figure 3). This difference could be related
to the presence of the chloride ions in the PBS solution.
For both experiments (i.e., with or without PBS), the
aliphatic hydrogen atoms of the coordinated PB
showed at 1.76 ppm (α‐CH2), 1.34 ppm (γ‐CH2) and
0.97 ppm (β‐CH2), while the signals of the released PB
were detected at 2.53 ppm (α‐CH2), 2.09 ppm (γ‐CH2)
and 1.76 ppm (β‐CH2) (Figure 3 and supporting informa-
tion, Figure S6), correlating well with those of a free PB
anion measured on NaPB in the same mixture of solvents.
Note: The composition of the hydrolysis product (i.e., 1‐
OD2 or 1‐OH2 according the used solvents) of the studied
complexes, was proved by RP‐HPLC and 1H NMR
experiments (supporting information, Figures S4 and S5).

ŠTARHA ET AL.
7 of 11
3.3 | Interaction with reduced glutathione
3.4 | In vitro cytotoxicity
(GSH)
Complex 1‐PB is comparably cytotoxic against the
cisplatin‐sensitive A2780 cell line (IC50 = 15.8 1.8 μM)
and its variant bearing the acquired resistance to cisplatin
(A2780R) with the IC50 value of 13.0 1.4 μM. Intrigu-
ingly, this is not the case of precursor 1‐Cl, whose activity
against the A2780 cells was ca 2‐fold higher than against
the A2780R ones, thus being of the same trend as in the
case of cisplatin (supporting information, Table S1). The
The natural biomolecule GSH occurs abundantly in cells
including cancer ones,[16] which is why the interaction
of newly developed metallodrugs with GSH is worth
studying, and provides important information regarding
the fate of the studied compounds in cells. The results of
1
the time‐dependent H NMR experiments (20% MeOD‐
d4/80% D2O; pH = 4.7) proved that the PB ligand releases
from complex 1‐PB more rapidly in the presence of GSH
(2 molar equiv.) as compared with the above‐mentioned
experiments without this biomolecule. Only the signals
of the released PB anion were found after 30 min of stand-
ing at ambient temperature, with no further changes
observed even after 48 h (Figure 3). To avoid an effect of
pH on the PB ligand release from the structure of complex
1‐PB in the presence of an excess of GSH, the experiments
were performed in either PBS solution (pH = 7.2) or in the
20% MeOD‐d4/80% D2O medium whose pH was adjusted
by an addition of KOD to 7.2. Similarly to the experiments
performed at lower pH, a total release of the PB ligand
was detected for the experiments carried out at pH = 7.2
(Figure 3). However, the rate of the PB ligand release
was different for these three experiments. In particular,
no coordinated PB ligand was detected after 30 min (pH
= 4.7; see above), 2 h (pH = 7.2; with an addition of
KOD) and after 24 h in phosphate buffer.
relative activity (RA
=
IC50(cisplatin)/IC50(1‐PB))
calculated for complex 1‐PB against the reference drug
cisplatin (supporting information, Table S1) equalled ca
0.4 at the A2780 cells and ca 1.6 at the A2780R ones,
where complex 1‐PB exhibited higher activity than
conventional cisplatin. Thus, the resistance factor (RF =
IC50(A2780R)/IC50(A2780)) is markedly lower for
complex 1‐PB than for its analogue 1‐Cl, cisplatin
(Figure 4) and similar Ir(III) complex [Ir(η5‐Cpbph
)
(phen)Cl]PF6 (RF reported in ref.[6] as 1.9) containing
more extended Cp‐ring (i.e., biphenyl instead of phenyl
of Cpph) than complexes 1‐PB and 1‐Cl.
Generally said, the pair of the A2780 and A2780R cell
lines is widely used for evaluation of the ability of newly
developed anticancer agents (including the half‐sandwich
Ir(III) ones[6,13a,18]) to overcome the acquired resistance to
cisplatin. However, the A2780/A2780R model possibly
provides another advantage, because these two cell lines
are known to differ in the intracellular GSH level.[19] In
particular, the GSH level is higher in the resistant
A2780R cells than in the A2780 ones. High GSH level in
the resistant cancer cells (such as A2780R) is known to
be disadvantageous for both platinum[20] and non‐plati-
num[21] anticancer complexes. However, a GSH role was
very recently reported as positive for some so called
GSH‐activatable cytotoxic complexes, such as [Ru2(μ‐
abpy)(bphen)4]Cl4; abpy = 4,4``‐azobis(2,2`‐bipyridine),
For all three variants of the GSH interaction experi-
ments, new signals of the covalent adduct with GSH,
i.e., [Ir(η5‐Cpph)(phen)(SG)]+ (1‐SG), were detected at
3.52 ppm (Cys‐α CH) and 2.16 ppm (Cys‐β CH2)
(supporting information, Figure S7), which correlated
with the previously reported results of similar half‐sand-
wich Ir–SG adducts.[17] Formation of the 1‐SG species
was further proved by the RP‐HPLC/ESI+ (tR
=
11.18 min; supporting information, Figure S8) and ESI+
(found: 876.1 m/z, calc.: 876.2 m/z; supporting informa-
tion, Figure S9) mass spectra of the mixture of complex
1‐PB and GSH.
bphen
=
4,7‐diphenyl‐1,10‐phenanthroline.[22] With
respect to the observed different solution behaviour of
complex 1‐PB in the presence or absence of GSH, an
FIGURE 4 Resistance (RF; left) and selectivity (SF; right) factors, defined as RF = IC50(A2780R)/IC50(A2780), and SF = IC50(MRC‐5)/
IC50(A2780 or A2780R) calculated from the in vitro cytotoxicity results for complexes 1‐PB and 1‐Cl. Selectivity factors were calculated for
both the A2780 (left bars) and A2780 (right bars) cell lines. CDDP = cisplatin

8 of 11
ŠTARHA ET AL.
activating effect of GSH on complex 1‐PB has to be taken
into account for further studies of complex 1‐PB and
analogues.
Besides a relatively high cytotoxicity of complex 1‐PB
against A2780 and A2780R, its low toxicity towards
carcinoma cells induced higher Δψm depleted cell popula-
tions of the treated A2780 cells (83.2%), than observed for
the positive control (i.e., CCCP) and cisplatin (18.3%)
(Figure 5 and supporting information, Table S3). The
mitochondria depletion is usually connected with the
elevated level of ROS, affecting the redox equilibrium in
the treated cancer cells,[25] as reported also for the half‐
sandwich Ir(III) complexes.[6,7] For this reason, the ability
of complexes 1‐PB and 1‐Cl to induce ROS and/or super-
oxide (SO) in the A2780 cells was evaluated by the
standard assay. Both complexes 1‐PB and 1‐Cl compara-
bly induced the formation of the ROS and SO radicals in
the A2780 cells, whose majorities (93.6% and 82.3% cells,
respectively) were detected in the upper right quadrant
of the obtained dot‐plots (Figure 6 and supporting
information, Table S4).
the MRC‐5 human non‐cancerous cells (IC50
=
124.1 16.7μM)wasencouragingaswell;IC50 =7.2 3.2μM
for cisplatin. The selectivity factors (SF = IC50(MRC‐5)/
IC50(A2780 or A2780R)), are markedly higher for complex
1‐PB than for cisplatin (Figure 4), implying higher
selectivity of complex 1‐PB towards the cancer cells over
the normal ones, than for cisplatin. The cellular accumu-
lation by the A2780 cells was comparable for complexes
1‐PB (167.4 13.9 ng Ir/106 cells) and 1‐Cl (144.0 7.8 ng
Ir/106 cells).
3.5 | Flow cytometry cell‐based studies
3.6 | Inhibition of histone deacetylase
activity in cell‐free media
Complex 1‐PB induced different cell cycle alternation in
the A2780 cells than cisplatin (Figure 5 and supporting
information, Table S2), which is usually accepted as an
evidence for different mechanisms of biological action of
the studied agents (1‐PB vs. cisplatin in this case). For
the cells treated by complex 1‐PB, only slightly higher
populations than for the untreated cells were observed
in the S‐cell cycle phase, which was connected with lower
populations in the G0/G1‐ and G2/M‐cell cycle phases
observed for complex 1‐PB than for the negative control
(supporting information, Figure S10). Interestingly,
similar Ir(III) complexes, such as the apoptosis‐inducing
complex [Ir(η5‐Cpbph)(phen)Cl]PF6, were reported as
inducing a G0/G1‐cell cycle phase arrest.[5c,6]
Therapeutic action of the applied anticancer agents
usually alters the mitochondrial membrane potential
(Δψm),[23] as reported for half‐sandwich Ir(III) agents[6,7]
as well as for cisplatin.[24] The results of the used mito-
chondrial membrane potential assay indicated that the
application of complex 1‐PB to the A2780 human ovarian
The biological action of sodium 4‐phenylbutyrate (NaPB)
is connected with the inhibition of histone deacetylases.[9]
It has been recently reported that Pt(IV) complexes
containing releasable 4‐phenylbutyrato O‐donor ligand
exhibit higher HDAC inhibitory efficiency than analogues
without this ligand.[3b] Bearing these facts in mind, we
performed the HDAC activity cell‐free studies for complex
1‐PB. It has been observed that complex 1‐PB effectively
inhibited the HDAC activity to 66.6 3.0 % (Figure 7).
The reference compounds NaPB and cisplatin reduced
the HDAC activity to 74.5 3.2 %, and 90.6 4.1 %, respec-
tively. The obtained results imply that the HDAC inhibi-
tory efficiency of complex 1‐PB is induced not only by
the released PB ligand but by both the Ir‐containing
species and the released PB ligand. Similarly to the results
obtained for complex 1‐PB, a slightly higher HDAC
activity inhibitory efficiency was recently reported for
half‐sandwich Ru(II) (98.4% efficiency) and Rh(III)
FIGURE 5 Summary of the flow cytometry studies of the A2780 cell populations (%) in the cell cycle phases (PI/RNase staining); given as
arithmetic mean from three independent experiments (left), and the mitochondria membrane potential flow cytometry analysis depicted for
the untreated A2780 cells (control; green) and the A2780 cells treated by 1‐PB (red), cisplatin (pink) and CCCP (positive control; red) (right)

ŠTARHA ET AL.
9 of 11
FIGURE 6 The cell populations observed by flow cytometry studies of the induction of ROS and superoxide, for 1‐PB (light blue), and
negative (red) and positive (pyocyanin; green) controls (top), and the results of EPR spin trapping experiments ‐ (A) blank (DEPMPO/1‐
OH2), (B) blank (DEPMPO/H2O2), mixtures of DEPMPO and H2O2 with 1‐PB (C), 1‐Cl (D) or 1‐OH2 (E), (F) simulated spectrum (see text for
parameters), and (G) represents E immediately after the addition of GSH
(98.9% efficiency) complexes containing the SAHA‐like
derivative of 1,10‐phenanthroline, as compared with the
free ligand (95.1% efficiency); SAHA = a known HDAC
inhibitor suberoylanilide hydroxamic acid (99.5%
efficiency).[26]
3.7 | EPR spin‐trapping studies
It is known that various compounds induce the free radi-
cal formation in the presence of hydrogen peroxide.[27]
The formed free radicals can interact with various spin
traps (e.g., DEPMPO or DMPO) allowing the EPR spin
trapping (EPR‐ST) studies of such adducts. However, to
FIGURE
7
The results of the HDAC activity evaluation
the best of our knowledge, similar EPR‐ST studies have
not been reported for half‐sandwich iridium(III) com-
plexes, known to induce the ROS in vitro in cancer cells.
The EPR spin‐trapping experiments were performed,
using the DEPMPO spin trap and hydrogen peroxide as
a substrate.[27] Only a very weak characteristic signal of
performed using a Fluorometric HDAC Activity Assay Kit. The
symbols (+) and (‐) represent the presence and absence,
respectively, of the corresponding compound. The depicted data
represent percentage of HDAC activity relative to a positive control
(no inhibitor). Data are shown as the mean SD of three
independent experiments

10 of 11
ŠTARHA ET AL.
the DEPMPO adduct with the hydroxyl radicals was
detected in the samples containing complex 1‐PB or its
precursor 1‐Cl (Figure 6). On the other hand, the [Ir(η5‐
Cpph)(phen)(H2O)]2+(1‐OH2) species, representing the
hydrolysis product of both complexes 1‐PB and 1‐Cl,
readily reacted with H2O2 resulting in the formation of
the EPR‐active adduct of DEPMPO with the hydroxyl rad-
icals.[28] Addition of 2 mol equiv. of GSH quenched the
observed signal. The EPR spectrum of the trapped radical
was fitted using EasySpin[29] with the following parame-
ters: g = 2.006, A(N) = 38.7 MHz, A(P) = 133 MHz and
A(H) = 35.2 MHz (Figure 6). Similar results (g = 2.01,
A(N) = 39.4 MHz, A(P) = 132.3 MHz and A(H) =
37.2 MHz) have been very recently reported for a series
of half‐sandwich Os(II) complexes, where the hydroxido
species (i.e., hydrolysate) showed higher efficiency of the
·OH generation than the co‐studied iodido and chlorido
complexes.[30]
her help with the syntheses, Ms. Kateřina Kubešová for
performing the cytotoxicity experiments, Dr. Pawel Jewula
for his help with the flow cytometry experiments, Mrs.
Pavla Richterová for performing elemental analysis and
Dr. Alena Klanicová for recording the FTIR spectra.
ORCID
Pavel Štarha http://orcid.org/0000-0003-0422-045X
Zdeněk Trávníček http://orcid.org/0000-0002-5890-7874
Zdeněk Dvořák http://orcid.org/0000-0002-3938-3585
REFERENCES
[1] L. Kelland, Nat. Rev. Cancer 2007, 7, 573.
[2] a) E. Gabano, M. Ravera, D. Osella, Dalton Trans. 2014, 43,
9813. b) A. K. Renfrew, Metallomics 2014, 6, 1324.
[3] a) S. Dhar, S. J. Lippard, Proc. Natl. Acad. Sci. U. S. A. 2009, 106,
22199. b) R. Raveendran, J. P. Braude, E. Wexselblatt, V.
Novohradsky, O. Stuchlikova, V. Brabec, V. Gandin, D. Gibson,
Chem. Sci. 2016, 7, 2381.
4 | CONCLUSIONS
[4] a) S. Medici, M. Peana, V. M. Nurchi, J. I. Lachowicz, G.
Crisponi, M. A. Zoroddu, Coord. Chem. Rev. 2015, 284, 329. b)
N. P. E. Barry, P. J. Sadler, Chem. Commun. 2013, 49, 5106. c)
R. Trondol, P. Heffeter, C. R. Kowol, M. A. Jakupec, W. Berger,
B. K. Keppler, Chem. Sci. 2014, 5, 2925.
In conclusion, 1‐PB represents a novel biologically active
iridium complex combining the potent half‐sandwich
Ir(III) species with the O‐donor releasable HDAC inhibi-
tor 4‐phenylbutyrate (PB). Complex 1‐PB hydrolysed in
the water‐containing solution, resulting in the release of
PB. The rate and extent of the PB ligand release is higher
in the PBS buffer as well as in the presence of the intracel-
lular thiol GSH. Complex 1‐PB exhibited comparable
activity against both the used variants of the ovarian
carcinoma cells (A2780, A2780R). The difference in a
cytotoxic profile of 1‐PB (RF = 0.82) and 1‐Cl (RF =
1.76) could be connected with a positive (activating) role
of GSH in the mechanism of action of 1‐PB. The results
of flow cytometric studies suggested that the application
of 1‐PB to the A2780 cells led to the different cell cycle
changes as well as to higher mitochondria membrane
depolarization, as compared with cisplatin. Further,
complex 1‐PB effectively inhibited the HDAC activity.
The results of the EPR spin trapping studies indicated that
hydrolysis could be a key mechanistic step, because only
the aqua complex 1‐OH2 induced hydroxyl radical, which
was not the case of complexes 1‐PB and 1‐Cl. Altogether,
the results of this case study can be helpful for further
development and research of half‐sandwich Ir(III)
complexes, which behave as prodrugs releasing two cyto-
toxic, mechanistically independent species.
[5] a) Z. Liu, A. Habtemariam, A. M. Pizarro, S. A. Fletcher, A.
Kisova, O. Vrana, L. Salassa, P. C. A. Bruijnincx, G. J. Clarkson,
V. Brabec, P. J. Sadler, J. Med. Chem. 2011, 54, 3011. b) C. H.
Leung, H. J. Zhong, D. S. H. Chan, D. L. Ma, Coord. Chem.
Rev. 2013, 257, 1764. c) Z. Liu, P. J. Sadler, Acc. Chem. Res.
2014, 47, 1174.
[6] V. Novohradsky, L. Zerzankova, J. Stepankova, A. Kisova, H.
Kostrhunova, Z. Liu, P. J. Sadler, J. Kasparkova, V. Brabec,
Metallomics 2014, 6, 1491.
[7] Z. Liu, I. Romero‐Canelón, A. Habtemariam, G. Clarkson, P. J.
Sadler, Organometallics 2014, 33, 5324.
[8] P. Štarha, A. Habtemariam, I. Romero‐Canelón, G. J. Clarkson,
P. J. Sadler, Inorg. Chem. 2016, 55, 2324.
[9] a) T. Iannitti, B. Palmieri, Drugs R. D. 2011, 11, 227. b) P. A.
Marks, Expert Opin. Invest. Drugs 2010, 19, 1049.
[10] D. M. Morris, M. McGeagh, D. De Peña, J. S. Merola, Polyhe-
dron 2014, 84, 120.
[11] P. Štarha, Z. Trávníček, R. Křikavová, Z. Dvořák, Molecules
2016, 21, 1725.
[12] H. E. Gottlieb, V. Kotlyar, A. Nudelman, J. Org. Chem. 1997, 62,
7512.
[13] a) R. Pettinari, F. Marchetti, C. Pettinari, F. Condello, A.
Petrini, R. Scopelliti, T. Riedel, P. J. Dyson, Dalton Trans.
2015, 44, 20523. b) A. Casini, F. Edafe, M. Erlandsson, L.
Gonsalvi, A. Ciancetta, N. Re, A. Ienco, L. Messori, M.
Peruzzini, P. J. Dyson, Dalton Trans. 2010, 39, 5556.
ACKNOWLEDGEMENTS
We thank the projects LO1305, CZ.1.05/2.1.00/19.0377 and
[14] N. Gulyaeva, A. Zaslavsky, P. Lechner, M. Chlenov, A. Chait, B.
17‐08512Y for financial support, Ms. Lucie Hanousková for
Zaslavsky, Eur. J. Pharm. Sci. 2002, 17, 81.

ŠTARHA ET AL.
11 of 11
[15] a) E. Bencze, B. V. Lokshin, J. Mink, W. A. Herrmann, F. E.
Kühn, J. Organomet. Chem. 2001, 627, 55. b) P. Govindaswamy,
Y. A. Mozharivskyj, M. R. Kollipara, Polyhedron 2005, 24, 1710.
c) H. G. Brittain, Cryst. Growth Des. 2010, 10, 1990.
Thomas, A. Berger, J. S. Carew, Clin. Cancer Res. 2013, 19,
3577. b) S. Alas, C. P. Ng, B. Bonavida, Clin. Cancer Res. 2002,
8, 836.
[25] M. L. Circu, T. Y. Aw, Free Radical Biol. Med. 2010, 48, 749.
[16] D. Montero, C. Tachibana, J. R. Winther, C. Appenzeller‐
Herzog, Redox Biol. 2013, 1, 508.
[26] J. M. Cross, T. R. Blower, N. Gallagher, J. H. Gill, K. L. Rockley,
J. W. Walton, ChemPlusChem 2016, 81, 1276.
[17] Z. Liu, I. Romero‐Canelón, B. Qamar, J. M. Hearn, A.
Habtemariam, N. P. E. Barry, A. M. Pizarro, G. J. Clarkson, P.
J. Sadler, Angew. Chem. Int. Ed. 2014, 53, 3941.
[27] a) T. Ozawa, A. Hanaki, J. Chem. Soc., Chem. Commun. 1991,
330. b) I. Yamazaki, L. H. Piette, J. Biol. Chem. 1990, 265, 13589.
[28] C. Frejaville, H. Karoui, B. Tuccio, F. Le Moigne, M. Culcasi, S.
[18] a) J. Yellol, S. A. Pérez, A. Buceta, G. Yellol, A. Donaire, P.
Szumlas, P. J. Bednarski, G. Makhloufi, C. Janiak, A. Espinosa,
J. Ruiz, J. Med. Chem. 2015, 58, 7310. b) R. M. Lord, A. J.
Hebden, C. M. Pask, I. R. Henderson, S. J. Allison, S. L.
Shepherd, R. M. Phillips, P. C. McGowan, J. Med. Chem. 2015,
58, 4940.
Pietri, R. Lauricella, P. Tordo, J. Med. Chem. 1995, 38, 258.
[29] S. Stoll, A. Schweiger, J. Magn. Reson. 2006, 178, 42.
[30] R. J. Needham, C. Sanchez‐Cano, X. Zhang, I. Romero‐
Canelón, A. Habtemariam, M. S. Cooper, L. Meszaros, G. J.
Clarkson, P. J. Blower, P. J. Sadler, Angew. Chem. Int. Ed.
2017, 56, 1017.
[19] a) A. K. Godwin, A. Meister, P. J. O'Dwyer, C. S. Huang, T. C.
Hamilton, M. E. Anderson, Proc. Natl. Acad. Sci. U. S. A.
1992, 89, 3070. b) B. C. Bohrens, T. C. Hamilton, H. Masuda,
K. R. Grotzinger, J. Whang‐Peng, K. G. Louie, T. Knutsen, W.
M. McKoy, R. C. Young, R. F. Ozols, Cancer Res. 1987, 47, 414.
SUPPORTING INFORMATION
Additional Supporting Information may be found online
in the supporting information tab for this article.
[20] P. Mistry, L. R. Kelland, G. Abel, S. Sidhar, K. R. Harrap, Br. J.
Cancer 1991, 64, 215.
[21] I. Romero‐Canelón, M. Mos, P. J. Sadler, J. Med. Chem. 2015,
58, 7874.
How to cite this article: Štarha P, Trávníček Z,
Drahoš B, Herchel R, Dvořák Z. Cell‐based studies
of the first‐in‐class half‐sandwich Ir(III) complex
containing histone deacetylase inhibitor 4‐
phenylbutyrate. Appl Organometal Chem. 2018;
e4246. https://doi.org/10.1002/aoc.4246
[22] L. Zeng, S. Kuang, G. Li, C. Jin, L. Ji, H. Chao, Chem. Commun.
2017, 53, 1977.
[23] V. Ribas, C. García‐Ruiz, J. C. Fernández‐Checa, Front.
Pharmacol. 2014, 5, 151.
[24] a) S. T. Nawrocki, K. R. Kelly, P. G. Smith, C. M. Espitia, A.
Possemato, S. A. Beausoleil, M. Milhollen, S. Blakemore, M.