Tagging (arene)ruthenium(II) anticancer complexes with fluorescent labels

Article abstract of DOI:10.1002/ejic.200700144

Fluorescent (arene)ruthenium(II) complexes have been prepared by tagging a small fluorogenic reporter onto the chelating ligand of complexes of the type [(η⁶-arene)RuCl(Z)]⁺ (Z = chelating ligand). Complexes [(η⁶-p-cym)-Ru

Full text of DOI:10.1002/ejic.200700144

FULL PAPER
DOI: 10.1002/ejic.200700144
Tagging (Arene)ruthenium(II) Anticancer Complexes with Fluorescent Labels
Fabio Zobi,[a] Beeta Balali Mood,[a] Peter A. Wood,[a] Francesca P. A. Fabbiani,[a]
Simon Parsons,[a] and Peter J. Sadler*[a]
Keywords: Anticancer agents / Organometallic compounds / Fluorescence / Ruthenium / Hydrolysis
Fluorescent (arene)ruthenium(II) complexes have been pre-
pared by tagging a small fluorogenic reporter onto the che-
lating ligand of complexes of the type [(η6-arene)RuCl(Z)]+
ily hydrolyze to form the aqua complexes. These are stable
at acidic pH forming 10–15% of the corresponding hydroxido
complexes in buffered solution (25 m
M HEPES) as the pH is
(Z
=
chelating
ligand).
Complexes
[(η6-p-cym)-
raised to a physiological value (pH = 7.44). Under these con-
ditions, 4 (but not 2 or 3) undergoes a fast pH-dependent
reversible intramolecular rearrangement. Experimental data
and semiempirical calculations indicate that the major spe-
cies arising from this transformation is a complex with a tri-
dentate chelating ligand following deprotonation at the ni-
trogen atom of the amide group. Esterase-catalyzed hydroly-
sis of 3 liberates isatoic acid (MIAH) and generates 2 indicat-
ing that the complex is a substrate for the enzyme. Com-
plexes similar to 3 may have potential for esterase-activated
Ru-based prodrug delivery systems.
RuCl(NNO)](Cl) (2), [(η6-p-cym)RuCl(L3)](Cl) (3) and [(η6-p-
cym)RuCl(L4)](Cl) (4) {p-cym = p-cymene, NNO = 2-[(2-ami-
noethyl)amino]ethanol, L3 = 2-[(2-aminoethyl)amino]ethyl-2-
(methylamino)benzoate and L4
= N-{2-[(2-aminoethyl)-
amino]ethyl}-2-(methylamino)benzamide} were obtained in
good yield from the reaction of the Ru dimer [(η6-p-cym)-
RuCl2]2 (1) and the corresponding ligand. The compounds
have been fully characterized and their X-ray crystal struc-
tures are reported. Compounds 3 and 4 show a photolumi-
nescence response centered at 435 nm with partial fluores-
cence quenching of the fluorogenic reporters L3 and L4 upon
coordination to the metal center. Species 2–4 show good solu-
bility both in water and organic solvents. In water, 2–4 read-
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
biphenyl)Ru(en)Cl][PF6] is as cytotoxic to cancer cells as
carboplatin.[1] Although the cellular target for (arene)ruthe-
nium compounds and the mechanism underlying their bio-
logical effects are not yet known, much of the research has
focused on the interaction of these complexes with DNA
bases and oligonucleotides.[1,5–10] In the same report Morris
et al. described strong and selective binding to G bases on
DNA, while no inhibition of topoisomerase I or II was de-
tected. Subsequently, Chen et al. showed that [(η6-arene)-
Ru(en)]2+ complexes exhibit a remarkably high preference
for binding to guanine N7.[5,6] Such binding is stabilized by
H-bonding between the O6 of G and NH of the en ligand
and strong arene–nucleobase stacking when the arene is ex-
tended, suggesting a potential DNA-binding mode, involv-
ing simultaneous arene intercalation and RuII coordina-
tion.[5,6] Hydrophobic interactions were described both in
solution and in the solid state, and it was suggested that it
is the flexibility of the arene ligand that allows for simulta-
neous arene–base stacking and N7-covalent binding. The
remarkable site selectivity of the complexes was further
demonstrated by Novakova et al. who showed by transcrip-
tion mapping experiments that in cell-free media the inter-
action of [(η6-arene)Ru(en)]2+ complexes with DNA occurs
at G N7.[11] In the same report noncovalent, hydrophobic
interaction between the arene ligand and DNA was demon-
strated by competitive ethidium displacement.
Introduction
Organometallic ruthenium(II) complexes of the type
[(η6-arene)RuCl(X)(Y)] (where X, Y are monodentate or
chelating ligands) comprise a relatively new group of anti-
cancer compounds which are cytotoxic to cancer cells, in-
cluding cisplatin-resistant cell lines.[1,2] The geometry of the
complexes is generally described as pseudo-octahedral
“three-legged piano-stool” in which a π-bonded arene li-
gand provides the seat supported by a σ-bonded chlorido
and a chelating ligand which form the three legs of the
stool. The arene ligand is strongly bound and stabilizes the
ruthenium(II) oxidation state. It resists hydrolysis and pro-
vides a lipophilic side to the complex.[3,4] Most of the halide
compounds, however, are ionic and exhibit good aqueous
solubility and appear to undergo rapid aquation (substitu-
tion of bound Cl by H2O).[4] Thus, while η6-arene–RuII
bonds are inert toward hydrolysis, the monofunctional com-
plexes [(η6-arene)RuII(en)(Cl)]+ (en = ethylenediamine)
readily lose their chlorido ligand and transform into the
corresponding, more reactive, aquated species.[4–6]
This type of compound has continued to attract atten-
tion since it was reported by Morris et al. in 2001 that [(η6-
[a] School of Chemistry, University of Edinburgh,
West Mains Road, Edinburgh EH9 3JJ, U.K.
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2783

P. J. Sadler et al.
FULL PAPER
Reactions of the (arene)chlorido(ethylenediamine)RuII
Owing to these considerations, we have synthesized
complexes with nucleotides appear to involve initial aqua- methylisatoic anhydride (MIA) based fluorescent deriva-
tion of the arene complexes, and reactions of the aqua tives of (arene)ruthenium complexes and studied their prop-
complexes with cyclic guanosine monophosphate are faster erties. The MIA probe was chosen for its relatively small
than those of the chlorido complexes.[4–6] Wang et al. size, a feature that we anticipated might not alter signifi-
showed that rates of aquation of these (arene)ruthenium an- cantly the properties of the complexes. The probe has found
ticancer agents are Ͼ20ϫ faster than that of cisplatin. The wide applications in biochemistry, and it has been used to
aqua complexes behave as weak acids with pKa values for label a variety of biological substrates ranging from RNA
the aqua adducts ranging from 7.7 to 8.[4]
and Ras-like low-molecular-weight GTP-binding proteins
In contrast to cis-[PtCl2(NH3)2], cisplatin, [(η6-arene)- to relatively large enzymes like kinases and ATPases.[26–30]
Ru(en)]2+ complexes show little binding to adenine. This We report here the synthesis and characterization of MIA-
is attributable to repulsive interactions between the exo- based fluorescent derivatives of (arene)ruthenium com-
cyclic amino group of the base and (en)Ru NH2 plexes (Scheme 1). We have also studied the interaction of
groups.[5,6] Replacing the neutral en by anionic acetyl complexes 3 and 4 (Scheme 1) with porcine liver esterase
acetonate as the chelating ligand changes the nucleobase (PLE), and we show that esterase-catalyzed hydrolysis reac-
specificity, and the overall affinity for adenosine was tions can liberate methylisatoic acid (MIAH) from complex
demonstrated to be greater than for guanosine.[7] The 3 suggesting a possible use of similar derivatives in esterase-
cytotoxicity of the complexes increases with the size of activated Ru-based prodrug delivery systems.
the arene ring system in the series arene = benzene Ͻ p-
cymene
Ͻ biphenyl Ͻ dihydroanthracene Ͻ tetra-
hydroanthracene which further supports a potential
DNA-binding mode, involving simultaneous arene inter-
calation and RuII coordination.[2]
Other cytotoxic (arene)Ru complexes of general formula
[(η6-arene)RuCl2(X)] (X = monodentate ligand) have also
recently been reported, including 1,3,5-triaza-7-phospha-
adamantane (PTA) and imidazole compounds described by
Dyson et al.[12–16] These complexes show activity towards
the TS/A mouse adenocarcinoma cancer cell line and no
apparent cytotoxicity towards HBL-100 human mammary
(non-tumor) cell line.[13,14] The cellular target and the
mechanism underlying the biological effects for these (ar-
ene)ruthenium compounds are also not yet known. The
molecules, however, exhibit a pH-dependent DNA damage
such that at pH values typical of hypoxic cells DNA is dam-
aged, whereas at pH values characteristic of healthy cells
little or no damage is detected.[13,14] The complexes de-
scribed by Dyson and co-workers may act by a different
mechanism from those of general formula [(η6-arene)-
RuCl(X)(Y)]. Loss of the arene is observed in the former
complexes, and it has been suggested that single-stranded
DNA may wrap around the metal center with formation of
multiple coordinative N-donor bonds.[12]
Scheme 1. Structure and abbreviations of ligands, reagents and
metal complexes.
In view of the importance of understanding the bio-
logical distribution of (arene)ruthenium compounds, and
thus elucidating their mechanism of action, we have tagged
them with a small fluorescent probe. Fluorescence micro-
scopy has been widely used to study cellular distribution of
organic drugs,[17–19] and recently it has been applied suc-
cessfully to platinum complexes.[20–24] In the latter case,
Results and Discussion
Synthetic Aspects and General Considerations
Reaction of the dimer [(η6-p-cym)RuCl2]2 (1) with
fluorogenic reporters of different size have been used. These 2.2 mol-equiv. of the corresponding ligand afforded the
range from large carboxyfluorescein diacetate,[20–22] to me- complexes [(η6-p-cym)RuCl(NNO)](Cl) (2), [(η6-p-cym)-
dium-size diaminoanthracene[23,24] to small dinitrophenyl RuCl(L3)](Cl) (3) and [(η6-p-cym)RuCl(L4)](Cl) (4) in good
derivatives.[20–22] Cellular uptake of labeled compounds is yields (Scheme 2). Complex 2 was first prepared for a direct
affected by large fluorophores, and it has been suggested conjugation reaction with MIA. This strategy, however,
that fluorescein-labeled compounds diffuse in and out of proved unsuccessful, and we were able to detect only small
the cells and intracellular compartments at a greater rate amounts of 3 by this route. We then proceeded by the or-
than the parent compounds.[25]
ganic synthesis of ligands (Scheme 2), and we used 2 as a
2784
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
standard for comparison with species 3 and 4. The com- complexes 2, 4 and 3 are shown in Figures 1, 2 and 3,
1
pounds were characterized by H NMR spectroscopy and respectively. All complexes have the familiar pseudo-octahe-
electrospray ionization mass spectrometry (ESI-MS). The dral “three-legged pianostool” geometry as described pre-
¯
ESI mass spectra of 2–4 in water, methanol or CH3CN pro- viously. Complex 2 crystallizes in space group P1 with one
vided parent peaks corresponding to the cations. De- solvent molecule of methanol and shows disorder in the
pending on the cone voltage, peaks corresponding to fur- methyl groups of the isopropyl substituent of p-cym. The
ther loss of HCl were also observed. All spectra confirmed disorder was modelled by allowing ½ occupancy for each
the presence of the expected products.
of the four carbon atoms involved. The Cl– counter anion
bridges the hydroxy groups of the solvent molecule and the
pendent arm of the chelating ligand through two hydrogen
bonds.
Table 1. Selected bond lengths [Å] and angles [°] for cations of 2,
3 and 4.[a]
2
3
4
Ru–Cl
Ru–NЈ
Ru–NЈЈ
NЈ–Ru–NЈЈ
NЈ–Ru–Cl
NЈЈ–Ru–Cl
2.4257(10)
2.125(4)
2.168(3)
79.48(14)
83.89(10)
86.92(9)
2.4148(8)
2.132(3)
2.159(3)
80.32(10)
82.71(8)
86.88(8)
2.4149(18)
2.115(6)
2.142(6)
79.7(2)
85.13(17)
87.83(17)
[a] NЈ and NЈЈ refer to the primary and secondary amine, respec-
tively, of the chelate ligand.
Scheme 2. Synthesis of MIA-related ligands and complexes. Condi-
tions: i: MIA, CH3CN, K2CO3, room temp. ii: TFA, CH2Cl2, room
temp. iii: MIA, CHCl3, K2CO3, room temp. iv: CHCl3, room temp.
Complexes 2–4 showed good solubility both in water and
organic solvents. The NMR spectra of compounds 3 and 4
in deuteriated chloroform contain peaks between δ = 6.5
and 8.1 ppm that correspond to the aromatic protons of the
2-(methylamino)benzoate/benzamide fragment; aliphatic
proton signals are observed between δ = 2.5 and 4.7 ppm.
Peaks for arene ligand protons are observed between δ =
5.4 and 5.9 ppm, with the CH(CH3)2 proton displaying a
characteristic septet between δ = 1.7 and 2 ppm. Methyl
group signals appear at δ ≈ 1.3 ppm. In water, the com-
pounds readily form the corresponding aqua complexes.
These are stable at acidic pH*, and no hydroxido species
could be detected under these conditions. The pKa values
of the aqua complexes also lie above 7 in agreement with
values reported for similar arene species.[4,7] As the pH is
raised to within the physiological range (buffered solutions,
pH = 7.44) the complexes behave as weak acids forming
about 10–15% of the corresponding hydroxido species.
Complexes 2 and 3 are stable under these conditions for
hours, but 4 slowly undergoes an intramolecular rearrange-
ment following deprotonation at the amide nitrogen atom.
This process is described in detail in the following sections.
Attempts to obtain X-ray quality crystals of the aqua and
hydroxido complexes were unsuccessful, but the chlorido
adducts were readily crystallized from organic solvents.
Figure 1. ORTEP diagram and atom numbering scheme for the
cation of 2 (50% probability ellipsoids). The H atoms have been
omitted for clarity.
Figure 2. ORTEP diagram and atom numbering scheme for the
cation of 4 (50% probability ellipsoids). The H atoms have been
omitted for clarity. Two independent molecules of 4 crystallize in
X-ray Crystallography
Crystallographic data are listed in Table 4 and selected
bond lengths and angles in Table 1. The structures of the the same unit cell; only one is shown.
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2785

P. J. Sadler et al.
FULL PAPER
Figure 4. Schematic representation and definition of the angle α.
counterion shows disorder in one oxygen atom. In this case
also the disorder was modelled by allowing ½ occupancy
Figure 3. ORTEP diagram and atom numbering scheme for the
cation of 3 (50% probability ellipsoids). The H atoms have been for each oxygen atom. In this structure the plane defined
omitted for clarity.
by the 2-(methylamino)benzoate ring is almost perpendicu-
lar to that of the p-cym ring (α = 88°, Figure 4).
¯
Complex 4 also crystallized in space group P1 with two
For structures of the cations of 2, 3 and 4, metal–nitro-
independent molecules in the asymmetric unit. These differ gen and metal–chlorido bonds all fall within normal values
by the relative orientation of the 2-(methylamino)benz- (Table 1).[1–7] Bond lengths and distances around the metal
amide ring with respect to the aromatic arene ring (see Fig- coordination sphere are similar (within 3σ) in all three
ure 4). In one molecule the plane defined by the six carbon structures indicating that addition of 2-(methylamino)ben-
atoms of the 2-(methylamino)benzamide ring forms an an- zoate/benzamide has little effect on the overall coordination
–
gle α of 28° with the plane of p-cym, while in the other of the bidentate chelate. In the crystal structure of the ClO4
molecule the benzamide ring is rotated with an angle α of salt of 3, short van der Waals contacts are found between
47°. Three solvent molecules of chloroform co-crystallize p-cym hydrogen atoms and the carbonyl oxygen atom and
with the complex with two Cl– counterions completing the the CH3NH group of the 2-(methylamino)benzoate frag-
content of the asymmetric unit. The counterions form H- ment of adjacent molecules. These are shown in Figure 5.
bonds with NH and NH2 groups of the chelating ligand.
Such short contacts involving coordinated arenes are be-
In order to obtain crystals of 3, the Cl– counterion had coming increasingly recognized.[31] They have also been ob-
to be removed by addition of AgClO4. Thus, complex 3, as served in [(η6-p-cym)M(acac)Cl] (M = Ru or Os) com-
–
a ClO4 salt, crystallized in space group P21/n. No solvent plexes,[7] and it has been suggested that the weak polariza-
–
molecules are present in the asymmetric unit. The ClO4
tion of arene ring C–H bonds may lead to interactions be-
Figure 5. Short contacts involving arene hydrogen atoms of the cation in the crystal structure of 3.
2786
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
tween the coordinated arene and H-bond acceptors in bio- for hours under these conditions. Formation of 4c is ac-
logical target sites such as DNA or RNA.[32] It is worth companied by the appearance of a fourth species (4d in Fig-
noting that we found no evidence of π-stacking interactions ure 6). The relative integrated intensity of the signals corre-
in any of the reported structures.
sponding to 4d account for about 2% of the total p-cym
Complexes 3 and 4 are the first examples of metal com- resonances compared to 9% of the hydroxido species 4b
plexes with a pendent 2-(methylamino)benzoate/benzamide and 89% of 4c. We tentatively assign these 4d resonances
ring, thus restricting structural comparisons with similar to an HEPES complex of 4. Although sulfonates are weak
compounds. To the best of our knowledge, the only other ligands, the assignment is reasonable given the relatively
instance in which the same fragments have been used in high concentration of the buffer. Moreover, it has been
coordination chemistry is in the preparation of Bosnich’s shown that under similar conditions, reactions of (arene)-
type of ligands in complexes investigated for cooperative ruthenium species with nucleotides produce ca. 1.5% of
dimetallic redox reactivity. In those complexes, however, the stable phosphate-bound complexes.[6] It is worth noting that
fragment is part of a larger dinucleating ligand frame- the relative concentration of the hydroxido species 4b re-
work.[33–38]
mains unchanged during the course of the transformation
indicating that the aqua complex 4a is the reactive species.
Relative to peaks for 4a, the NHu proton signals suffer
a dramatic downfield shift for 4c. This indicates either a
Stability in Aqueous Buffered Solutions
In water the compounds readily form the corresponding strong interaction of these hydrogen atoms with nucleo-
aqua complexes which are stable at acidic pH. In a buffered philic atoms or a strong deshielding influence of the aro-
solution (25 m HEPES, pH = 7.44) the complexes behave matic ring current of the benzamide fragment; or a combi-
as weak acids forming about 10–15% of the corresponding nation of both. Related [(η6-arene)Ru(en)guanine]+ com-
hydroxido species. Complexes 2 and 3 are stable under these plexes, for example, show stereospecific en NHd bonding
conditions for hours, while 4 slowly undergoes an intramo- with guanine O6, and these species in solution show a
lecular rearrangement following deprotonation at the amide downfield shift of the chelate en NHd hydrogen signals of
nitrogen atom as shown in Scheme 3.
up to 1 ppm compared to the parent aqua complex.[5,6] It
1
The H NMR spectrum of 4 in the buffered solution is is unlikely that the observed shifts result from a partial
shown in Figure 6. After ca. 15 min, three sets of p-cym overlap or π–π stacking of the aromatic rings, but it is pos-
signals were visible in the spectrum of the complex (Fig- sible that the shifts are also a consequence of the ring cur-
ure 6a). The major species is the aqua complex 4a with p- rent generated by the orientation of the benzamide ring.
cym resonances at δ = 5.52, 5.67 and 5.72 ppm. The corre- Proton signals of the benzamide fragment show an unusual
sponding resonances of the hydroxido species 4b are shifted behaviour. The four non-equivalent resonances collapse
upfield to δ = 5.18 and 5.38 ppm. A third set of p-cym sig- into two signals, which might suggest an increased sym-
nals assigned to 4c appears at δ = 5.15 and 5.30 ppm. For- metry of the ring. This, however, is not possible as it would
mation of 4c was clearly evident from the time-dependent imply a complete rearrangement of the ring substituents.
evolution of the NMR spectrum. The reaction was com- Therefore, we assign this to a coincidental overlap of reso-
plete within 3 h, and the mixture thus obtained was stable nances.
Scheme 3. Hydrolysis of 4 in buffered solution (25 m HEPES, pH = 7.44).
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2787

P. J. Sadler et al.
FULL PAPER
1
Figure 6. H NMR spectrum (δ = 5.1–7.9 ppm) of a 3 m solution of [(η6-p-cym)Ru(L4)Cl]Cl (4) in HEPES buffer (25 m, pH = 7.44).
(a) After ca. 15 min; (b–d) 30-min time intervals; (e) after 3 h.
Substitution reactions of the aquated species 4a are likely Ru(H2O)3]2+, this electronic effect makes RuII behave like
to follow the associative interchange mechanism previously an RuIII center in arene species,[39] thus making substitution
described for related complexes. Wang et al. showed that reactions appear to occur by associative interchange mecha-
aquation and anation substitution reactions of (arene)Ru nisms.
species are influenced by steric hindrance of the arene li-
Changes observed in the NMR spectrum are ac-
gand and that the rates of aquation of the halide complexes companied by a transition in the UV/Vis spectrum as
follow the order Cl ≈ Br Ͼ I, consistent with an associative shown in Figure 7. Soon after dissolution of 4 in the buffer
mechanism.[4] Substituents on the p-cym ligand exert only solution (pH = 7.44), the UV/Vis spectrum showed a broad
minor steric effects when compared to the complexes d-d transition between 300 and 360 nm. Within 2 h, a rela-
studied by Wang et al.; however, as a π-acid ligand, p-cym tively sharp peak at 294 nm replaced this absorption indi-
depletes electron density on the metal center making it cating that H2O is replaced by a ligand which causes a
more electropositive. As demonstrated for [(η6-C6H6)- greater splitting of the metal d-orbitals. The intramolecular
Figure 7. Left: Time evolution of the UV/Vis spectrum of a 0.5 m solution of 4 (270–430 nm, 25 m HEPES buffer, pH = 7.44, 5-min
time intervals). Right: UV/Vis spectra of 4 at fixed pH values.
2788
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
rearrangement of 4a is a pH-dependent reversible process basis set. Structures are shown in Figure 8 and selected
(Figure 7). Aliquots of a stock solution of 4 in water were bond lengths are listed in Table 2. Structure B was chosen
added to 25 m HEPES solutions, and the pH was adjusted as a model of 4c, while C (as a 4d model) was optimized
to different values by addition of HCl. The mixtures were for comparison reasons. The tridentate chelate in B is stabi-
allowed to equilibrate at room temperature for 24 h, and lized by hydrogen bonds involving the 2-(methylamino)-
their UV/Vis and ESI-MS spectra were recorded. UV/Vis benzamide ring and the Hd proton of the primary amine
analysis indicated that 4a is stable at pH Ͻ 3 (Figure 7). coordinated to the metal center. Compared to A, Ru–N dis-
Species 4c began to form at pH ≈ 3.5 and became the pre- tances increase slightly in B upon coordination of the de-
dominant species at higher pH values.
protonated amide nitrogen atom. These results reflect an
ESI-MS analysis confirmed this observation (Supporting increased electron density at the metal center that weakens
Information: Figure S1). At pH Ͻ 3.5 only the parent peak the σ-donor bonds of the chelate. The opposite effect is
corresponding to the cation of
4
is observed calculated for C, where Ru–N distances decrease (on
(C22H34ClN4ORu: m/z = 507.15, matching the predicted average by 0.14 Å) by allowing HEPES to enter the Ru co-
isotope pattern). At higher pH values a second peak corre- ordination sphere and occupy a coordination site through
sponding to loss of HCl from the parent cation of 4 was the sulfonate oxygen atom.
also observed. This peak became the predominant signal at
pH Ͼ 4.
Table 2. Summary of selected INDO/1-optimized bond lengths
[pm].
Geometry Optimization and INDO/S Analysis and
Ru–NH2
Ru–NH
Ru–X[a]
Electronic Spectra
Structure A
Structure B
Structure C
213
216
200
215
220
199
241
199
194
The molecular structure of 4c was also confirmed by
theoretical calculations. Molecular structures B and C were
initially modeled from the crystal structure of the cation of
4 (structure A) and were optimized at the semiempirical
[a] X refers to the atom occupying the exchangeable coordination
site. Thus, for structure A, X = Cl; for structure B, X = N; for
level using default Hyperchem 7.5 INDO/1 functional and structure C, X = O.
Figure 8. INDO/1-optimized geometry of structure A (4), B and C and relative predicted (gas-phase, INDO/S) spectra. Vertical bars are
the predicted oscillator strengths and locations of individual computed transitions.
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2789

P. J. Sadler et al.
FULL PAPER
In both B and C the planarity of the arene ligand is ergies are less affected by distortion. Consequently, it is ex-
somewhat lost compared to A. Both the methyl and the pected that the UV spectra would show their bands shifted
isopropyl group of p-cymene are slightly bent out of the bathochromically with respect to the normal, unstrained
plane defined by the six carbon atoms of the p-cym aro- compounds of comparable structure and degree of substitu-
matic ring. The distortion of the arene is uniform and rela- tion.
tively contained in C with both methyl and isopropyl
With respect to the d-d transition, the experimental UV/
groups deviating from planarity by about 0.06 Å. In B the Vis spectrum of 4a at pH = 3 (Figure 7) shows a broad
distortion is more pronounced and uneven with the methyl absorption spanning from about 300 to 360 nm. As the pH
and the isopropyl substituents out of the benzene plane by is raised to 5.1, the transition is blueshifted with λmax
≈
0.14 and 0.06 Å, respectively. In A the observed deviation 294 nm. The predicted spectrum of structure A shows five
is on average 0.04 Å. The net effect of these distortions is peaks ranging from 285 to 370 nm in close agreement with
clearly visible in the predicted electronic spectra of the com- the experimental spectrum of 4a at low pH. The same d-d
plexes discussed below.
transition suffers two opposite shifts in the calculated spec-
Predicted absorptions of structures A, B and C are listed tra of B and C. A hypsochromic shift in the absorption is
in Table 3, and spectra are shown in Figure 8. Calculations predicted for B (peaks in the range of 266–311 nm) which
for the electronic spectra are essentially gas-phase, but the is consistent with the replacement of H2O or Cl– by the
Zerner INDO parameters were obtained from solution- deprotonated amide group with stronger σ-donor character.
phase experimental data and are more likely to predict spec- For C, a bathochromic shift is predicted (peaks in the range
tra where the solvent may play an important role.[40–43] In of 320–357 nm). Of the two optimized structures, B shows
all cases calculated spectra show relatively strong transi- the closest similarity to the UV/Vis spectrum of 4a at pH
tions in the 220–235 nm region and a series of peaks of = 5.1 (see Figures 8 and 9). Therefore, in agreement with
lesser intensity at longer wavelength. Overall the spectra experimental data, we argue that in 25 m HEPES at pH
compare well with experimental data. Complexes of the = 7.44, 4a undergoes deprotonation at the nitrogen atom of
type [(η6-arene)M(X)(Y)] (M = RuII or OsII; X, Y = simple
σ-donors, monodentate or chelating ligands) show a first
band at 300–400 nm which is assigned to a metal d-d transi-
tion. A second band appears at around 250 nm with a
shoulder at about 225 nm, and these are assigned to a
d(M)Ǟπ*(X/Y) and to an arene πǞπ* transition, respec-
tively.[44] The calculated spectra replicate well these features.
The loss of planarity of the arene ligand in B and C is
reflected by a bathochromic shift of the πǞπ* transition
relative to A (Figure 8). The degree of shift in energy can
be related to the degree of arene distortion in the predicted
structures. The shift is more pronounced in B (∆ = +9 nm)
then in C (∆ = +5 nm). These predictions are reasonable
because, while most steric effects in organic compounds
cause steric inhibition of resonance which determines hyp-
Figure 9. Experimental (black) and predicted (gas-phase, INDO/S)
spectra of 4 (dark grey) and of structure B (light grey with squares).
The vertical bars are the predicted oscillator strengths and loca-
tions of individual computed transitions. The oscillator strengths
are arbitrarily scaled to match the overall spectra; they retain their
sochromic displacements in the electronic spectrum, geome-
try distortion tend to cause the opposite effect.[45,46] Distor-
tion of an aromatic system raises the energy of both ground
and excited states of the molecule. However, due to the
greater antibonding character of the excited states their en- correct relative intensity.
Table 3. Predicted spectra for structures A (cation of 4), B and C.[a]
Structure A
Structure B
Structure C
Wavenumber
Oscillator strength
Wavenumber
Oscillator strength
Wavenumber
Oscillator strength
(ϫ103 cm–1)
(f)
(ϫ103 cm–1)
(f)
(ϫ103 cm–1)
(f)
27.0
30.9
32.1
32.8
35.1
38.9
42.9
45.1
45.7
47.2
0.07
0.06
0.09
0.14
0.11
0.04
0.1
1.09
0.49
0.05
32.1
34.3
37.6
41.3
42.3
43.3
44.1
0.09
0.06
0.105
0.045
0.15
0.885
0.285
28.0
30.0
30.4
31.3
36.9
44.3
44.8
45.9
46.7
0.036
0.188
0.064
0.1
0.04
0.828
0.248
0.032
0.072
[a] Only bands for allowed transitions (f Ͼ 0) are listed. Energies are given to one decimal place since this is only a model system.
2790 Eur. J. Inorg. Chem. 2007, 2783–2796
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
the amide group generating a complex with a tridentate che-
late encompassing the RuII center as depicted by structure
B.
Fluorescence Spectroscopy and PLE-Catalyzed Hydrolysis
In aqueous solutions (pH Ͻ 7) compounds 3 and 4 show
a photoluminescence response upon excitation at 325 nm.
The emission spectra of both compounds show a broad
peak centered at around 435 nm. As expected, quenching
of fluorescence is observed in the spectra of 3 and 4 com-
pared to the spectra of methylisatoic acid (MIAH) and the
corresponding free ligands L3 and L4 (Figure 10; Support-
ing Information: Figures S2 and S3). This is due to an in-
ternal heavy-atom effect. The quenching of fluorescence by
heavy atoms is a phenomenon first identified as “physical
quenching” by F. Perrin,[47,48] and it has been the subject of
many studies.[49] Heavy atoms in the fluorophore, or in close
contact with it, increase the rate of intersystem crossing
(ISC) by strengthening spin-orbit coupling,[50,51] as for elec-
tronic transitions of heavy atoms. Thus, the decrease of
fluorescence yield (radiative transition S1ǞS0) is explained
by an increase in the probability of the competing S1ǞTn
radiationless transition of the fluorophore.
Figure 11. pH-dependent fluorescence responses (λex = 325 nm, λem
= 438 nm) of a 7 µ unbuffered aqueous solution of 3 (circles) and
of 4 (triangles).
formed in 25 m HEPES-buffered solutions at pH = 7.44.
Samples of the metal complexes were prepared in the buffer
and allowed to equilibrate at room temperature for 5 h. Un-
der these conditions, 3 hydrolyses to form the correspond-
ing aqua (3a), hydroxido (3b) and HEPES (3d) complexes
with relative percentage abundances of ca. 84, 14 and 2%,
respectively. Complex 4, on the other hand, is mainly pres-
ent as the tridentate chelate 4c following deprotonation at
the amide nitrogen atom as described above. After the
equilibration period, the enzyme was added and the interac-
tions monitored by 1H NMR and fluorescence spec-
troscopy.
Treatment of 3 with PLE led to an increase in the photo-
luminescence intensity in its emission spectrum ac-
companied by a blue shift in the λem maximum from 438 to
398 nm as shown in Figure 12. The reaction was complete
after ca. 150 h of incubation, when no further spectral
change could be detected. The hypsochromic shift clearly
indicates hydrolysis of the L3 ester bond and release of free
MIAH. No change was observed in the fluorescence spec-
trum of 4 over the same time period. Under our experimen-
Figure 10. Fluorescence responses (λex = 325 nm) of a 7 µ aque-
ous solution of 4 (a), and of MIAH (b).
Complexes 3 and 4 show pH-dependent fluorescence re-
sponses (Figure 11) with maxima centered at around neu-
tral pH. Complex 3 shows a local photoluminescence maxi-
mum at pH = 6; however, the fluorescence of the sample
increases exponentially at pH Ͼ 8 due to base-catalyzed
hydrolysis of the L3 ester bond. Under these conditions,
the conjugate base of MIAH is liberated, and the overall
fluorescence of the sample increases. This observation
prompted us to study the interaction of compounds 3 and
4 with porcine liver esterase (PLE) in order to determine if
the complexes are substrates for this enzyme.
Ester hydrolysis by cytosolic esterases is a common strat-
egy employed to trap photoluminescent organic sensors
(most often fluorescein-type sensors) within the cell or to
activate organic-based progrugs. Recently, the same strategy
has being considered for possible esterase-activated metal-
Figure 12. Fluorescence responses (λex = 325 nm) of a 10 µ solu-
tion of 3 (25 m HEPES, pH = 7.44) after addition of PLE (7.5 U,
based prodrug delivery systems.[52] Experiments were per- incubation at 298 K).
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2791

P. J. Sadler et al.
FULL PAPER
1
Figure 13. Low-field p-cym region of the H NMR spectra of a 2 m solution of 3 in HEPES buffer (25 m, pH = 7.44); (a) 5 h after
dissolution at 298 K, (b) after 1 week incubation with PLE (15 U) at 298 K, (c) p-cym peaks of 2 in HEPES buffer (15 min). Asterisks
indicate overlapping resonances of 3b and 3d.
tal conditions, PLE-catalyzed hydrolysis of the ester bond chelating ligand). The compounds show interesting photo-
in 3 appears to be slow (t1/2 ≈ 40 h) when compared to flu- luminescent properties and may find some application in
orophore-bearing organic molecules.[53] However, we are fluorescence microscopy studies aimed at elucidation of
not aware of comparable examples involving inorganic their cellular trafficking and thus provide new insights into
complexes of the type we have investigated. Platinum com- the mechanism of action of this new class of anticancer
plexes prepared by Reedijk and co-workers are all derivat- compounds. Our results show that a fluorogenic reporter
ized with fluorogenic reporters through amide bonds.[20–24] may be tagged onto these complexes either through an ester
Their chemistry has been concentrated on reactions with or an amide bond. If the fluorophore is linked through an
guanine for comparison with the parent compounds, and ester bond, the resulting compounds may be substrates for
they have been studied almost exclusively to elucidate their cytosolic esterases. In this case, esterase-catalyzed hydrolysis
cellular distribution. Their interactions with cytosolic ester- of the reporter could potentially alter the fluorescent profile
ases do not appear to have been investigated.
and consequently the resulting (apparent) cellular distribu-
In order to determine the fate of the ruthenium complex tion of the compounds. Derivatization of the compounds
in the PLE-catalyzed hydrolysis of 3, a millimolar sample through an ester bond, however, might prove useful in the
of the compound was incubated with the enzyme (HEPES design of esterase-activated Ru-based prodrug delivery sys-
buffer, D2O solution, pH = 7.44) for 7 d. At the end of the tems. This is a common strategy in medicinal chemistry, as
incubation period, a white precipitate was detected and it is the case, for example, with capecitabine, an orally avail-
1
later identified as a 2-(methylamino)benzoate salt. The H able prodrug of 5-deoxy-5-fluorouridine, which has recently
NMR spectrum of the remaining solution was recorded and come into clinical use for the treatment of breast and colo-
compared to the spectra of 2 and 3 in the same buffer. rectal cancers.[54–57] A similar approach focussed on plati-
These are shown in Figure 13. The spectrum of 3 in the num drugs was recently suggested by Lippard.[49] Finally,
buffer revealed the presence of three species: the aqua com- tagging a reporter (or a prodrug) to (arene)RuII complexes
plex (3a) as the major adduct, together with the hydroxido through an amide linkage is also promising; however, a
(3b) and HEPES (3d) adducts with overlapping resonances. spacer of proper length (Ͼ –[CH2]2–) must be employed.
The species distribution remained unaltered over a period Our study indicates that if a spacer of only two carbon
of 5 h. After 1 week, a new single major set of signals was atoms is used, then the amide nitrogen atom readily un-
visible (Figure 13b). The peaks show the same resonances dergoes deprotonation under physiological conditions, pro-
(δ = 5.38, 5.41 and 5.56 ppm) as for the p-cym peaks of 2 moting the formation of a stable five-membered chelate
in HEPES (Figure 13c). The experimental evidence clearly ring.
indicates that esterase-catalyzed hydrolysis of 3 liberates
MIAH and generates complex 2. Thus, although the hydrol-
ysis reaction appears to be slow, the results indicate that 3
is a substrate for the enzyme.
Experimental Section
Materials: Reagents were purchased from Aldrich and used with-
out further purification. A porcine liver esterase (PLE) suspension
was purchased from Fluka and used as supplied. [(η6-p-cym)-
RuCl2]2 (1) was prepared according to a published procedure.[58,59]
Conclusions
We have synthesized and characterized fluorescent (ar-
X-ray Crystallography: All diffraction data were collected using a
ene)RuII complexes of the type [(η6-arene)RuCl(Z)] (Z = Bruker Smart Apex CCD diffractometer equipped with an Oxford
2792
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
Table 4. Crystal structure data for complexes 2, 3 and 4.
2
3
4
Empirical formula
Formula mass
Crystal system
Crystal size [mm]
Space group
V [Å3]
a [Å]
b [Å]
c [Å]
α [°]
C15H30Cl2N2O2Ru
442.39
triclinic
C22H33Cl2N3O6Ru
607.50
monoclinic
0.22ϫ0.14ϫ0.08
P21/n
2530.47(12)
10.7840(3)
10.5940(3)
22.1650(5)
90
92.1530(10)
90
C23.50H35.50Cl6.50N4ORu
721.55
triclinic
0.80ϫ0.33ϫ0.15
0.75ϫ0.15ϫ0.15
¯
¯
P1
P1
930.18(14)
9.0664(8)
9.8934(8)
11.1677(9)
109.405(5)
90.206(6)
99.447(6)
150
3035.5(2)
13.9097(6)
15.4667(7)
15.7873(7)
80.444(3)
69.698(3)
72.812(3)
150
β [°]
γ [°]
T [K]
150
Z
2
4
4
R [F Ͼ 4σ(F)][a]
0.0498
0.0365
0.0804
0.1754
1.318
+1.170/–1.219
[b]
Rw
0.1384
0.0884
GOF[c]
0.7221[d]
+1.38/–2.08
0.8676[d]
+0.98/–0.74
[e]
Max/min ∆ρ [eÅ–3]
2 2
2 2
[a] R = ∑||Fo| – |Fc||/∑|Fo|. [b] Rw = [∑w(Fo2 – Fc ) /∑wFo
]
2 1/2. [c] GOF = [∑w(Fo2 – Fc ) /(n – p)]1/2 (n = number of reflections, p = number
of parameters). F Ͼ 4σ(F). [d] Refinement based on F with ca. 5300 data. [e] Refinement based on F with 12274 data.
Cryosystems low-temperature device operating at 150 K (Table 4).
Absorption corrections for all data sets were performed with the
multiscan procedure SADABS.[60] Structures were solved by direct
methods (SHELXS or SIR92).[61,62] Complexes were refined
water, and ethanol before use. Purified water was obtained from
a Millipore Milli-Q water purification system. Stock solutions of
reagents in water were made up to concentrations of 2 m and kept
at 4 °C in 500-µL aliquots. Portions were diluted to the required
against F2 using CRYSTALS (2 and 3)[63] or SHELXL (4).[61] H- concentrations immediately prior to each experiment. Fluorescence
atoms were placed in idealised positions. In 2 the iPr group is rota-
tionally disordered about the pivot carbon atom (C82), and in 3
one O-atom of the perchlorate anion was modelled over two sites.
The crystal of 4 was twinned. The diffraction pattern was indexed
with two orientation matrices related by a twofold rotation about
the (01–1) reciprocal lattice direction (CELL_NOW);[64] reflections
from both domains were integrated simultaneously (SAINT)[65]
and used in the refinement. The twin scale factor was 0.2811(13).
CCDC-634687 (2), -634689 (3) and -634688 (4) contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
and absorption data were measured in water and in HEPES buffer
(25 m, pH = 7.4), prepared with Millipore water. Solutions were
stored in clean, dry propylene containers and were filtered before
data acquisition. Fluorescence spectra were recorded from 350 to
600 nm. All measurements were performed in triplicate. Fluores-
cence spectroscopy measurements were performed with Perkin–El-
mer LS50B or Edinburgh Instruments FS900 Fluororimeters. Exci-
tation was set to 325 nm and emission was collected through a 350-
nm cut-off filter.
Computational Methods: Molecular structures were initially mod-
elled from the crystal structure of the cation of 4 and were opti-
mized at the semiempirical INDO/1 level. A tight self-consistent-
field convergence criterion (10–6 hartree) was employed in all calcu-
lations. The electronic spectra of the complexes were calculated
with the INDO/S method in the Zerner implementation[40–42,66]
using the Hyperchem (version 7.5) software.[43] The ruthenium and
nitrogen atomic parameters of Krogh-Jespersen et al. were used.[67]
The overlap weighting factors, fσ and fπ, for INDO/S calculations
were set at 1.267 and 0.585, respectively, and the number of CISs
in the CIS calculations was 1250 (25 occupied and 25 unoccupied
orbitals). The optical spectra of these ruthenium complexes were
calculated using the SWizard program.[68,69] Half-bandwidths ∆1/2
were assumed to be equal to 3000 cm–1, which is a typical half-
bandwidth value for octahedral ruthenium complexes.[70]
NMR Spectroscopy: 1H NMR spectra were acquired with Bruker
DMX 500 (1H, 500 MHz) or DPX 360 (1H, 360 MHz) spectrome-
ters using TBI probe-heads equipped with z-field gradients. Typi-
1
cally, H NMR spectra in D2O were acquired with water suppres-
1
sion by Shaka’s method or with presaturation. H NMR chemical
shifts were referenced internally to 1,4-dioxane (δ =3.75 ppm) for
aqueous solutions, and CHCl3 (δ =7.26 ppm) for [D]chloroform
solutions. All data processing was carried out using XWIN-NMR
version 2.0 (Bruker U.K. Ltd.).
Mass Spectrometry: Electrospray ionization mass spectra (ESIMS)
were obtained with a Micromass Platform II mass spectrometer,
and solutions were infused directly. The capillary voltage was 3.5 V,
and the cone voltage was varied between 10 and 45 V, depending
on sensitivity. The source temperature was 353 K. Mass spectra
were recorded in a scan range of m/z = 200–1200 for positive ions.
Synthetic Procedures
[(η6-p-cym)RuCl(NNO)](Cl) (2): [(η6-p-cym)RuCl2]2 (1) (100 mg,
0.16 mmol) was dissolved in CHCl3 (5 mL). To this solution
2.2 mol-equiv. of 2-[(2-aminoethyl)amino]ethanol (NNO) was
added and the mixture stirred for 12 h. A change in colour of the
solution from deep red to pale yellow was observed within 10 s of
addition of the ligand to 1. The yellow precipitate was then col-
lected by filtration, washed with ice-cold CHCl3 and then three
times with 5 mL of ice-cold diethyl ether. Yield: quantitative.
C14H26Cl2N2ORu (410.75): calcd. C 40.98, H 6.39, N 6.83; found
C 40.82, H 6.09, N 6.57. 1H NMR (500 MHz, CDCl3): δ = 1.56
pH Measurements: pH measurements were carried out using a
Corning 240 pH meter equipped with an Aldrich micro-combina-
tion electrode calibrated with Aldrich standard buffer solutions of
pH = 4, 7, and 10. For NMR samples prepared in D2O, no correc-
tion has been applied for the effect of deuterium on the glass elec-
trode.
Spectroscopic Measurements and Fluorometric Determinations: All
glassware was washed sequentially with 20% HNO3, deionised
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2793

P. J. Sadler et al.
FULL PAPER
(d, 6 H, cymene), 2.43 (s, 3 H, cymene), 2.08 (sept, 1 H, cymene),
N-{2-[(2-Aminoethyl)amino]ethyl}-2-(methylamino)benzamide (L4):
2.99 (t, 2 H, CH2), 3.18 (t, 2 H, CH2), 3.49 (t, 2 H, CH2), 3.56 (br., Typically, N-(2-aminoethyl)ethane-1,2-diamine (100 mg) was dis-
1 H, NH2), 4.06 (t, 2 H, CH2), 5.04 (br., 1 H, NH2), 6.63 (br., 1 solved in CHCl3 (2 mL). To this solution, MIA (100 mg) in CHCl3
H, OH), 7.33 (br., 1 H,NH) ppm. ESI MS (+ve): m/z = 375 [M]+.
Crystals suitable for X-ray diffraction were obtained by slow dif-
fusion of diethyl ether into a methanolic solution of the complex
(typically, 5 mg of complex in 250 µL of solvent; diffusion occurred
at 4 °C).
(2 mL) was added in 0.25-mL portions. Immediately after the ad-
dition of MIA, a white precipitate appeared. The reaction was al-
lowed to proceed for 2 h. The white precipitate was filtered off,
while the CHCl3-soluble fraction was concentrated to about 3 mL
and then loaded onto a preparative TLC plate. With 15% CH3OH
in CH2Cl2, all the impurities contained in the starting material
moved with the front, molecules derivatized with two or three MIA
moved to the middle of the plate, while L4 moved only a few cm
from the base line. Yield: 38 mg (32%). C12H20N4O (236.31): calcd.
C 60.99, H 8.53, N 23.71; found C 60.23, H 8.40, N 23.69. 1H
NMR (500 MHz, CDCl3): δ = 1.75 (br., NH2), 2.68 (m, 2 H, CH2),
2.79 (m, 2 H, CH2), 2.83 (m, 5 H, CH2 + CH3), 3.47 (m, 2 H,
CH2), 6.56 (t, 1 H, aromatic), 6.64 (d, 1 H, aromatic), 6.82 (br.,
NH), 7.30 (t, 1 H, aromatic), 7.36 (2, 1 H, aromatic), 7.47 (br.,
NH) ppm. ESI MS (+ve): m/z = 259.15 [M + Na+].
N,NЈ-Bis(tert-butoxycarbonyl)-2-[(2-aminoethyl)amino]ethanol (L1):[71]
To a solution of 2-[(2-aminoethyl)amino]ethanol (50 mmol) in
THF/H2O (1:1) (300 mL), KHCO3 (150 mmol) and tert-butoxycar-
bonyl anhydride (Boc2O) (120 mmol) were added consecutively at
0 °C. After the reaction mixture had reached room temp., the solu-
tion was stirred overnight. The mixture was then acidified to pH =
4 by careful addition of acetic acid at 0 °C and then extracted with
CH2Cl2 (4ϫ200 mL). The combined organic phases were dried
(Na2SO4) and concentrated under reduced pressure to give L1 as
a viscous colourless oil which crystallized on standing. L1 was used
for the next step without further purification. Yield: 6 g (60%). 1H
NMR (360 MHz, CDCl3): δ = 1.44 (s, 9 H, CH3), 1.47 (s, 9 H,
CH3), 3.32 (m, 2 H, CH2), 3.37 (m, 4 H, CH2), 3.74 (m, 2 H, CH2)
ppm. ESI MS (+ve): m/z = 343.37 [M + K+].
[(η6-p-cym)RuCl(L3)](Cl) (3): Ligand L3, (0.032 g, 0.13 mmol), was
dissolved in CH3OH/CHCl3 (20:80) (2.4 mL). This was added
dropwise to
a
solution of [(η6-p-cym)RuCl2]2 (1) (0.036 g,
0.059 mmol) in CHCl3 (2.8 mL) to prevent any initial precipitation
of the ligand. The colour of dimer solution changed from deep red
to orange once all of L3 was added. The mixture was stirred at
room temp. for 40 h. During this time, a yellow precipitate devel-
oped on the sides of reaction flask, while the colour of the reaction
mixture remained unchanged. Further precipitation was obtained
by addition of diethyl ether (10 mL). The yellow precipitate was
dried under vacuum. Yield: 22 mg (30%). C22H33Cl2N3O2Ru
(543.49): calcd. C 48.62, H 6.12, N 7.73; found C 47.98, H 6.07, N
7.91. 1H NMR (500 MHz, CDCl3): δ = 1.38 (d, 6 H, p-cymene),
1.74 (sept, 1 H, p-cymene), 2.46 (s, 3 H, p-cymene), 2.85–3.20 (m,
4 H, CH2), 3.21 (t, 2 H, CH2), 4.06 (t, 2 H, CH2), 4.72 (br., 1 H,
NH2), 5.67–5.78 (dd, 4 H, p-cymene), 6.14 (br., 1 H, NH2), 6.23
(br., 1 H, NH), 6.82 (t, 1 H, aromatic), 6.96 (d, 1 H, aromatic),
7.31 (br., 1 H, OH), 7.53 (t, 1 H, aromatic), 8.04 (d, 1 H, aromatic)
ppm. ESI MS (+ve): m/z = 508 [M]+. Crystals suitable for X-ray
diffraction were obtained by slow diffusion of pentane into a
CHCl3 solution of the complex (5 mg of complex + 2 mg of Na-
ClO4 in 250 µL of solvent; diffusion occurred at room temp.).
N,NЈ-Bis(tert-butoxycarbonyl)-2-[(2-aminoethyl)amino]ethyl-2-
(methylamino)benzoate (L1Ј): Boc-protected 2-[(2-aminoethyl)-
amino]ethanol (0.4 g, 1.39 mmol) was dissolved in dried CH3CN
(20 mL). N-Methylisatoic anhydride (MIA) (0.37 g, 2.08 mmol) was
added to this solution. An excess amount of K2CO3 (2 g,
27.8 mmol) was added to the above mixture. The mixture was
stirred at room temp. for 48 h, during which the reaction progress
was monitored by TLC. The desired product was separated from
the other possible byproducts and impurities by silica column
chromatography using a solvent system of (CH3OH/CH2Cl2,
12:88). The column was eluted with a total of 300 mL of eluent.
The fractions obtained were analyzed by 1H NMR spectroscopy,
and those containing the desired product were combined. The sol-
vents (CH3OH/CH2Cl2) were removed by rotary evaporation, and
the resulting yellow oil was stored at 4 °C for 18 h. During storage,
the product spontaneously crystallized into a light yellow solid.
Yield: 277 mg (45%). 1H NMR (500 MHz, CDCl3): δ = 1.40 (s, 18
H, CH3), 2.86 (s, 3 H, CH3), 3.26 (t, 2 H, CH2), 3.37 (t, 2 H, CH2),
3.58 (t, 2 H, CH2), 4.31 (t, 2 H, CH2), 4.92 (br., 1 H, NH), 5.15 [(η6-p-cym)RuCl(L4)](Cl) (4): 1 (44 mg) was dissolved in CHCl3
(br., 1 H, NH), 6.58 (t, 1 H, aromatic), 6.63 (d, 1 H, aromatic), (3 mL). To this solution 2.2 mol-equiv. of L4 (38 mg) in CHCl3
7.36 (t, 1 H, aromatic), 7.48 (d, 1 H, aromatic) ppm. ESI MS (+ve):
(1 mL) was added and the mixture stirred for 48 h. The colour of
the solution changed within 1 h from red to orange to olive-green.
The reaction was monitored by 1H NMR spectroscopy and
stopped when no further change could be observed. Diethyl ether
was added to the solution depositing a yellow precipitate which
was then collected by filtration and washed with diethyl ether.
Yield: 68 mg (78%). C22H34Cl2N4ORu (542.51): calcd. C 48.71, H
6.32, N 10.33; found C 48.08, H 5.97, N 9.98. 1H NMR (500 MHz,
CDCl3): δ = 1.22 (d, 6 H, CH3, p-cymene), 2.08 (sept, 1 H, p-
cymene), 2.24, (s, 3 H, CH3, p-cymene), 2.85, (s, 3 H, CH3), 3.7–
2.6, (m, 8 H, CH2), 5.44 (d, 1 H, p-cymene), 5.60 (d, 1 H, p-cy-
mene), 5.65 (d, 1 H, p-cymene), 5.91 (d, 1 H, p-cymene), 6.53 (t, 1
H, aromatic), 6.59 (d, 1 H, aromatic), 7.29 (t, 1 H, aromatic), 7.99
(d, 1 H, aromatic) ppm. ESI MS (+ve): m/z = 507.1 [M]+. Crystals
of 4 suitable for X-ray diffraction were obtained by slow diffusion
of pentane into a CHCl3 solution of the complex.
m/z = 439.1 [M]+, 338.1 [M – C5H10O2]+.
2-[(2-Aminoethyl)amino]ethyl-2-(methylamino)benzoate (L3): Boc
deprotection of L1Ј was attained by dropwise addition of 15 mol-
equiv. of TFA (1.28 mL, 3.45 mmol) to a solution of L1Ј (0.1 g,
0.23 mmol) in distilled CH2Cl2 (10 mL). The mixture was stirred at
room temp. for 48 h and was monitored by TLC. The solvents were
then removed by rotary evaporation. The crude product was re-
dissolved in methanol and purified with the aid of preparative thin
layer chromatography (TLC). The desired product was then ex-
tracted from preparative TLC silica by re-dissolution in methanol
and stirring at room temp. for 18 h. This solution was then filtered
under gravity, and all the solvents were removed from the filtrate
under vacuum. The product was a dark yellow oil. Yield: 28 mg
(51%). C12H19N3O2 (237.30): calcd. C 60.74, H 8.07, N 17.71;
found C 61.03, H 8.04, N 17.38. 1H NMR (500 MHz, D2O): δ =
2.06 (s, 1 H, CH3), 3.43 (t, 2 H, CH2), 3.53 (t, 2 H, CH2), 3.61 (t,
2 H, CH2), 4.58 (t, 2 H, CH2), 6.98 (t, 1 H, aromatic), 7.09 (d, 1
H, aromatic), 7.63 (t, 1 H, aromatic), 8.07 (d, 1 H, aromatic) ppm.
ESI MS (+ve): m/z = 238 [M – H]+.
Supporting Information (see footnote on the first page of this arti-
cle). Positive-ion ESI-MS spectra of a solution of 4 in 25 m
HEPES at different pH values; fluorescence response of L3 and L4
upon titration of 1.2 mol-equiv. of 1 (Figures S1–S3); ZINDO1/S
2794
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796

Fluorescent (Arene)ruthenium(II) Complexes
FULL PAPER
[26] K. A. Wilkinson, E. J. Merino, K. M. Weeks, J. Am. Chem.
Soc. 2005, 127, 4659–4667.
[27] T. K. Nomanbhoy, D. A. Leonard, D. Manor, R. A. Cerione,
Biochemistry 1996, 35, 4602–4608.
parameter files employed in the semiempirical calculations
(Table S1).
Acknowledgments
[28] J. P. Hutchinson, J. F. Eccleston, Biochemistry 2000, 39, 11348–
11359.
[29] A. Y. Jan, E. F. Johnson, A. J. Diamonti, K. L. Carraway III,
K. S. Anderson, Biochemistry 2000, 39, 9786–9803.
[30] S. P. Gilbert, M. L. Moyer, K. A. Johnson, Biochemistry 1998,
37, 792–799.
[31] D. Braga, F. Grepioni, G. R. Desiraju, J. Organomet. Chem.
1997, 548, 33–44.
We thank the Swiss National Science Foundation (SNF; fellowship
for F. Z.) for the support for this work, and Dr. Abraha Habtemar-
iam and Dr. Ana Maria Pizarro (Edinburgh) for helpful dis-
cussions.
[1] R. E. Morris, R. E. Aird, P. del S. Murdoch, H. Chen, J. Cum-
mings, N. D. Hughes, S. Parsons, A. Parkin, G. Boyd, D. I.
Jodrell, P. J. Sadler, J. Med. Chem. 2001, 44, 3616–3621.
[2] R. E. Aird, J. Cummings, A. A. Ritchie, M. Muir, R. E. Morris,
H. Chen, P. J. Sadler, D. I. Jodrell, Br. J. Cancer 2002, 86, 1652–
1657.
[3] L. Dadci, H. Elias, U. Frey, A. Hoernig, U. Koelle, A. E. Mer-
bach, H. Paulus, J. S. Schneider, Inorg. Chem. 1995, 34, 306–
315.
[4] F. Wang, H. Chen, S. Parsons, I. D. H. Oswald, J. E. Davidson,
P. J. Sadler, Chem. Eur. J. 2003, 9, 5810–5820.
[5] H. Chen, J. A. Parkinson, S. Parsons, R. A. Coxall, R. O.
Gould, P. J. Sadler, J. Am. Chem. Soc. 2002, 124, 3064–3082.
[6] H. Chen, J. A. Parkinson, R. E. Morris, P. J. Sadler, J. Am.
Chem. Soc. 2003, 125, 173–186.
[7] R. Fernández, M. Melchart, A. Habtemariam, S. Parsons, P. J.
Sadler, Chem. Eur. J. 2004, 10, 5173–5179.
[8] H.-K. Liu, F. Wang, J. A. Parkinson, J. Bella, P. J. Sadler,
Chem. Eur. J. 2006, 12, 6151–6165.
[9] H.-K. Liu, S. J. Berners-Price, F. Wang, J. A. Parkinson, J. Xu,
J. Bella, P. J. Sadler, Angew. Chem. Int. Ed. 2006, 45, 8153–
8156.
[10] F. Wang, J. Xu, A. Habtemariam, J. Bella, P. J. Sadler, J. Am.
Chem. Soc. 2005, 127, 17734–17743.
[11] O. Novakova, H. Chen, O. Vrana, A. Rodger, P. J. Sadler, V.
Brabec, Biochemistry 2003, 42, 11544–11554.
[12] A. Dorcier, P. J. Dyson, C. Gossens, U. Rothlisberger, R. Scop-
elliti, I. Tavernelli, Organometallics 2005, 24, 2114–2123.
[13] C. Scolaro, A. Bergamo, L. Brescacin, R. Delfino, M. Cocchi-
etto, G. Laurenczy, T. J. Geldbach, G. Sava, P. J. Dyson, J.
Med. Chem. 2005, 48, 4161–4171.
[14] C. Scolaro, T. J. Geldbach, S. Rochat, A. Dorcier, C. Gossens,
A. Bergamo, M. Cocchietto, I. Tavernelli, G. Sava, U. Rothlis-
berger, P. J. Dyson, Organometallics 2006, 25, 756–765.
[15] C. A. Vock, C. Scolaro, A. D. Phillips, R. Scopelliti, G. Sava,
P. J. Dyson, J. Med. Chem. 2006, 49, 5552–5561.
[16] W. H. Ang, E. Daldini, C. Scolaro, R. Scopelliti, L. Juillerat-
Jeannerat, P. J. Dyson, Inorg. Chem. 2006, 45, 9006–9013.
[17] G. Arancia, A. Calcabrini, S. Meschini, A. Molinari, Cytotech-
nology 1998, 27, 95–111.
[18] A. A. Hindenburg, J. E. Gervasoni Jr, S. Krishna, V. J. Stewart,
M. Rosado, J. Lutzky, K. Bhalla, M. A. Baker, R. N. Taub,
Cancer Res. 1989, 49, 4607–4614.
[19] A. K. Larsen, A. E. Escargueil, A. Skladanowski, Pharmacol.
Ther. 2000, 85, 217–229.
[20] G. V. Kalayda, G. Zhang, T. Abraham, H. J. Tanke, J. Reedijk,
J. Med. Chem. 2005; 48, 5191–5202.
[21] C. Molenaar, J. M. Teuben, R. J. Heetebrij, H. J. Tanke, J. Re-
edijk, J. Biol. Inorg. Chem. 2000, 5, 655–665.
[32] A. F. A. Peacock, A. Habtemariam, R. Fernandez, V. Walland,
F. P. A. Fabbiani, S. Parsons, R. E. Aird, D. I. Jodrell, P. J.
Sadler, J. Am. Chem. Soc. 2006, 128, 1739–1748.
[33] A. L. Gavrilova, B. Bosnich, Chem. Rev. 2004, 104, 349–383.
[34] A. L. Gavrilova, C. Jin Qin, R. D. Sommer, A. L. Rheingold,
B. Bosnich, J. Am. Chem. Soc. 2002, 124, 1714–1722.
[35] C. Incarvito, A. L. Rheingold, A. L. Gavrilova, C. Jin Qin, B.
Bosnich, Inorg. Chem. 2001, 40, 4101–4108.
[36] L. M. Liable-Sands, C. Incarvito, A. L. Rheingold, C. J. Qin,
A. L. Gavrilova, B. Bosnich, Inorg. Chem. 2001, 40, 2147–2155.
[37] C. Incarvito, A. L. Rheingold, C. J. Qin, A. L. Gavrilova, B.
Bosnich, Inorg. Chem. 2001, 40, 1386–1390.
[38] B. Bosnich, Inorg. Chem. 1999, 38, 2554–2562.
[39] M. Stebler-Rothlisberger, W. Hummel, P. A. Pittet, H. B. Burgi,
A. Ludi, A. E. Merbach, Inorg. Chem. 1988, 27, 1358–1363.
[40] J. Ridley, M. C. Zerner, Theor. Chim. Acta 1976, 42, 223–236.
[41] M. C. Zerner, G. H. Loew, R. F. Kirchner, U. T. Mueller-West-
erhoff, J. Am. Chem. Soc. 1980, 102, 589–599.
[42] W. P. Anderson, W. D. Edwards, M. C. Zerner, Inorg. Chem.
1986, 25, 2728–2732.
[43] S. I. Gorelsky, A. B. P. Lever, J. Organomet. Chem. 2001, 635,
187–196.
[44] Y. Hung, W.-J. Kung, H. Taube, Inorg. Chem. 1981, 20, 457–
463.
[45] J. Kulhánek, S. Böhm, K. Palát, O. Exner, J. Phys. Org. Chem.
2004, 17, 686–693.
[46] S. Böhm, O. Exner, Chem. Eur. J. 2000, 6, 3391–3398.
[47] F. Perrin, J. Phys. Radium 1926, 7, 390–401.
[48] F. Perrin, J. Chim. Phys. 1928, 25, 531.
[49] S. K. Lower, M. A. El-Sayed, Chem. Rev. 1966, 66, 199–241.
[50] D. S. McClure, J. Chem. Phys. 1949, 17, 905–913.
[51] M. Kasha, J. Chem. Phys. 1952, 20, 71–74.
[52] D. Wang, S. J. Lippard, Nature Rev. Drug Discovery 2005, 4,
307–320.
[53] C. C. Woodroofe, A. C. Won, S. J. Lippard, Inorg. Chem. 2005,
44, 3112–3120.
[54] F. Di Costanzo, A. Sdrobolini, S. Gasperoni, Crit. Rev. Oncol.
Hematol. 2000, 35, 101–108.
[55] J. L. Marshall, Oncology 2001, 15, 41–46.
[56] C. Twelves, M. Boyer, M. Findlay, J. Cassidy, C. Weitzel, C.
Barker, B. Osterwalder, C. Jamieson, K. Hieke, Eur. J. Cancer
2001, 37, 597–604.
[57] M. L. H. Wang, W. K. A. Yung, M. E. Royce, D. E. Schomer,
R. L. Theriault, D. F. Wogan, Am. J. Clin. Oncol. 2001, 24,
421–424.
[58] M. A. Bennett, A. K. Smith, J. Chem. Soc., Dalton Trans. 1974,
233–241.
[59] R. A. Zelonka, M. C. Baird, Can. J. Chem. 1972, 50, 3063–
3072.
[60] G. M. Sheldrick, University of Göttingen, Göttingen, Ger-
many, 2006.
[22] B. A. J. Jansen, P. Wielaard, G. V. Kalayda, M. Ferrari, C. Mo-
lenaar, H. J. Tanke, J. Brouwer, J. Reedijk, J. Biol. Inorg. Chem.
2004, 9, 403–413.
[23] G. V. Kalayda, B. A. J. Jansen, C. Molenaar, P. Wielaard, H. J.
Tanke, J. Reedijk, J. Biol. Inorg. Chem. 2004, 9, 414–422.
[24] G. V. Kalayda, B. A. J. Jansen, P. Wielaard, H. J. Tanke, J. Re-
edijk, J. Biol. Inorg. Chem. 2005, 10, 305–315.
[25] M. D. Hall, C. T. Dillon, M. Zhang, P. Beale, Z. Cai, B. Lai,
A. P. J. Stampfl, T. W. Hambley, J. Biol. Inorg. Chem. 2003, 8,
726–732.
[61]
G. M. Sheldrick, University of Göttingen, Göttingen, Ger-
many, 1997.
[62]
A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi,
M. C. Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr.
1994, 27, 435.
[63]
P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, D. J.
Watkin, J. Appl. Crystallogr. 2003, 36, 1487.
Eur. J. Inorg. Chem. 2007, 2783–2796
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2795

P. J. Sadler et al.
[69] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 270–283.
[70] A. B. P. Lever, Can. J. Chem. 2004, 82, 1102–1111.
[71] A different synthetic procedure was published: M. T. Schütte,
P. Schumacher, C. Unger, R. Mülhaupt, F. Kratz, Inorg. Chim.
Acta 1998, 267, 133–136.
FULL PAPER
[64] G. M. Sheldrick, University of Göttingen, Göttingen, Ger-
many, 2005.
[65] Bruker-Nonius, Bruker-AXS, Madison, Wisconsin, USA,
2006.
[66] J. Ridley, M. C. Zerner, Theor. Chim. Acta 1973, 32, 111–134.
[67] K. Krogh-Jespersen, J. D. Westbrook, J. A. Potenza, H. J. Schu-
gar, J. Am. Chem. Soc. 1987, 109, 7025–7031.
[68] S. I. Gorelsky, SWizard program. http://www.sg-chem.net/
swizard/.
Received: January 31, 2007
Published Online: May 3, 2007
2796
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 2783–2796