In vitro evaluation of rhodium and osmium RAPTA analogues: The case for organometallic anticancer drugs not based on ruthenium

Article abstract of DOI:10.1021/om060394o

Reaction of the dimer [(η⁵-C₅Me ₅)RhCl(μ₂-Cl)]₂ with 2 or 4 equiv of the water-soluble phosphine 1,3,5-triaza-7-phosphatricyclo[3.3.1.1]decane (pta) affords [Rh(η⁵-C₅Me₅

Full text of DOI:10.1021/om060394o

4090
Organometallics 2006, 25, 4090-4096
In Vitro Evaluation of Rhodium and Osmium RAPTA Analogues:
The Case for Organometallic Anticancer Drugs Not Based on
Ruthenium
Antoine Dorcier,† Wee Han Ang,† Sandra Bolan˜o,§ Luca Gonsalvi,§
Lucienne Juillerat-Jeannerat,‡ Ga`bor Laurenczy,† Maurizio Peruzzini,*,§
Andrew D. Phillips,†,§ Fabrizio Zanobini,§ and Paul J. Dyson*,†
Institut des Sciences et Inge´nierie Chimiques, Ecole Polytechnique Fe´de´rale de Lausanne (EPFL),
CH-1015 Lausanne, Switzerland, UniVersity Institute of Pathology, Centre Hospitalier UniVersitaire
Vaudois (CHUV), CH-1011 Lausanne, Switzerland, and Istituto di Chimica del Composti Organometallici,
Consiglio Nazionale delle Ricerche (ICCOM-CNR), Via Madonna del Piano 10, 50019 Sesto Fiorentino,
Firenze, Italy
ReceiVed May 9, 2006
Reaction of the dimer [(η5-C5Me5)RhCl(µ2-Cl)]2 with 2 or 4 equiv of the water-soluble phosphine
1,3,5-triaza-7-phosphatricyclo[3.3.1.1]decane (pta) affords [Rh(η5-C5Me5)(pta)Cl2] and [Rh(η5-C5-
Me5)(pta)2Cl]Cl, respectively. Both complexes have been characterized in solution by NMR spectroscopy
-
and in the solid state by single-crystal X-ray diffraction, the latter as the chloride and BPh4 salts. In
addition, the rhodium(I) complexes [Rh(η5-C5Me5)(CO)(pta)] and [Rh(η5-C5H5)(pta)2] have been prepared
from [Rh(η5-C5Me5)(CO)2] and [Rh(η5-C5H5)(PPh3)2], respectively, by reaction with pta. An in vitro
evaluation of these compounds, together with [Os(η6-C10H14)(pta)Cl2] and the well-characterized
antimetastasis drug [Ru(η6-C10H14)(pta)Cl2], RAPTA-C, was undertaken using HT29 colon carcinoma,
A549 lung carcinoma, and T47D breast carcinoma cells. In the HT29 cell line, the two nearest congeners
to [Ru(η6-C10H14)(pta)Cl2], viz., [Rh(η5-C5Me5)(pta)Cl2] and [Os(η6-C10H14)(pta)Cl2], demonstrated very
similar cytotoxicity profiles. [Rh(η5-C5Me5)(pta)Cl2] proved significantly more cytotoxic in A549 cells
and [Rh(η5-C5Me5)(pta)2Cl]Cl 3-fold more cytotoxic in T47D cells, both relative to RAPTA-C. These
data suggest that the development of organometallic anticancer drugs based on the neighboring elements
to ruthenium should not be overlooked.
transition metal cyclopentadienyl (metallocene) complexes, in
the 1970s.6 Although titanocene dichloride successfully com-
pleted phase II clinical trials, it has not gained clinical approval.7
We are not aware of any organometallic compound in clinical
trials for cancer, although ferrocifen, a ferrocenyl derivative of
tamoxifen,7 looks set to enter clinical trials in the near future,8
and related ferrocenyl derivatives are in clinical trials for
malaria.9 The only non-platinum transition metal compounds
currently in clinical trials are two coordination compounds based
on ruthenium, viz., [ImH][trans-RuCl4(DMSO)Im] (NAMI-A)10
and [ImH][trans-RuCl4Im2] (KP1019),11 which has stimulated
much interest in the medicinal properties of this metal. A gallium
compound is also in clinical trials, and main group compounds
have been extensively studied for their anticancer properties.12
Both the ruthenium compounds demonstrate a low general
Introduction
Following the success of cisplatin in the clinic since its
discovery in 1965,1 which remains the most widely used
anticancer drug to this day, employed in the treatment of
approximately 70% of all cancer patients, a large number of
other platinum-based drugs have been prepared and evaluated.2
Although cisplatin and platinum-based drugs, more generally,
are arguably the most successful class of anticancer drugs in
the world, they are not without problems; notably they exhibit
a high general toxicity, leading to undesirable side-effects,
although minimized by careful administration protocols, and
they are also inactive against certain types of cancers.3 Thus,
other metal coordination compounds have been evaluated in
cancer chemotherapy,4 and organometallic compounds have also
been evaluated extensively for their medicinal properties.5 The
earliest example of a highly successful organometallic anticancer
compound is titanocene dichloride, pioneered by Ko¨pf and Ko¨pf-
Maier, who investigated the antitumor activity of several early
(6) Koepf, H.; Koepf-Maier, P. Angew. Chem. 1979, 91, 509.
(7) Top, S.; Kaloun, E. B.; Vessieres, A.; Leclercq, G.; La¨ıos, I.;
Ourevitch, M.; Deuschel, C.; McGlinchey M. J.; Jaouen, G. ChemBioChem.
2003, 4, 754.
(8) Vessie`res, A.; Top, S.; Beck, W.; Hillard, E.; Jaouen, G. Dalton Trans.
2006, 529.
(9) Biot, C.; Glorian, G.; Maciejewski, L. A.; Brocard, J. S. J. Med.
Chem. 1997, 40, 3715.
(10) (a) Sava, G.; Capozzi, I.; Bergamo, A.; Gagliardi, R.; Cocchietto,
M.; Masiero, L.; Onisto, M.; Alessio, E.; Mestroni, G.; Garbisa, S. Int. J.
Cancer 1996, 68, 60. (b) Redemaker-Lakhai, J. M.; van den Bongard, D.;
Pluim, D.; Beijnen, J. H.; Schellens, J. H. M. Clin. Cancer Res. 2004, 10,
3717.
(11) (a) Galanski, M.; Arion, V. B.; Jakupec, M. A.; Keppler, B. K. Curr.
Pharm. Des. 2003, 9, 2078. (b) Reisner, E.; Arion, V. B.; Guedes da Silva,
M. F. C.; Lichtenecker, R.; Eichinger, A.; Keppler, B. K.; Kukushkin, V.
Yu.; Pombeiro, A. J. L. Inorg. Chem. 2004, 43, 7083.
* To whom correspondence should be addressed. E-mail: mperuzzini@
iccom.cnr.it; paul.dyson@epfl.ch.
† EPFL.
‡ CHUV.
§ ICCOM-CNR.
(1) Rosenberg, B.; Vancamp, L.; Krigas, T. Nature 1965, 205, 698.
(2) Wong, E.; Giandomenico, C. M. Chem. ReV. 1999, 99, 2451.
(3) Wang, D.; Lippard, S. J. Nat. ReV. Drug DiscoVery 2005, 4, 307.
(4) Clarke, M. J.; Zhu, F.; Frasca, D. R. Chem. ReV. 1999, 99, 2511.
(5) (a) Allardyce, C. S.; Dorcier, A.; Scolaro, C.; Dyson, P. J. Appl.
Organomet. Chem. 2005, 19, 1. (b) Jaouen, G., Ed. Bioorganometallics;
Wiley-VCH: Weinheim, 2005.
10.1021/om060394o CCC: $33.50 © 2006 American Chemical Society
Publication on Web 07/13/2006

Rhodium and Osmium RAPTA Analogues
Organometallics, Vol. 25, No. 17, 2006 4091
Chart 1
toxicity, which contrasts to the pharmacological properties of
platinum drugs, and this low toxicity has been ascribed to
selective uptake of ruthenium compounds by cancer cells
mediated via the transferrin cycle13 and to mechanisms such as
activation through reduction.14 Interest in inorganic ruthenium
drugs has extended to organometallic compounds, with a
growing body of work on ruthenium(II)-arene derivatives.15
For example, ruthenium(II)-arene compounds with imida-
zole,16 alanine- and guanine-derived co-ligands,17 ethylenedi-
amine,18 disulfoxide,19 and pta20 co-ligands have been evaluated
(pta ) 1,3,5-triaza-7-phosphatricyclo[3.3.1.1]decane).21 The pta-
derived compounds appear to impart a pH-dependent activity,
which is conducive to providing excellent pharmacological
properties. In this paper we describe closely related congeners
to the well-characterized (in vitro and in vivo) ruthenium(II)-
arene-pta compounds based on osmium and rhodium and show
that these metals are also worth considering within the context
of organometallic anticancer drugs.
Results and Discussion
The compounds used in the comparative in vitro study (see
below) are shown in Chart 1. [Ru(η6-C10H14)(pta)Cl2], RAP-
TA-C (1),20a and the osmium analogue 222 have been reported
previously, the former having been subjected to an in vivo study,
proving it to be a highly selective agent for the treatment of
secondary (metastasis) tumors with excellent pharmacokinetic
properties.23
Scheme 1
The rhodium compounds used in the study are new and were
prepared from the dimer [(η5-C5Me5)RhCl(µ2-Cl)]2 according
to the routes shown in Schemes 1 and 2. Direct reaction of the
rhodium dimer with 2 or 4 equiv of pta affords the highly water
soluble species [Rh(η5-C5Me5)(pta)Cl2] (3) and [Rh(η5-C5-
Me5)(pta)2Cl]Cl (4‚Cl), respectively. The chloride counteranion
in 4‚Cl is easily substituted by other anions, e.g., tetraphenylbo-
rate, by stirring with NaBPh4 in MeOH to give [Rh(η5-C5-
Me5)(pta)2Cl]BPh4 (4g‚BPh4). The electrospray ionization mass
spectrum (ESI-MS) of 3 in MeOH is dominated by a peak at
m/z ) 466, which corresponds to the protonated parent ion [Rh-
(12) (a) Jakupec, M. A.; Keppler, B. K. Curr. Top. Med. Chem. 2004,
4, 1575. (b) Jakupec M. A.; Keppler B. K. Met. Ions Biol. Syst. 2004, 42,
425.
(13) For example see: Kratz F.; Hartmann, M.; Keppler, B.; Messori,
L. J. Biol. Chem. 1994, 269, 2581.
Scheme 2
(14) For example see: Jakupec, M. A.; Reisner, E.; Eichinger, A.;
Pongratz, M.; Arion, V. B.; Galanski, M.; Hartinger, C. G.; Keppler, B. K.
J. Med. Chem. 2005, 48, 2831.
(15) For reviews see: (a) Yan, Y. K.; Melchart, M.; Habtemarian, A.;
Sadler, P. J. Chem. Commun. 2005, 4764. (b) Dyson, P. J.; Sava, G. Dalton
Trans. 2006, 1929.
(16) Dale, L.; Tocher, J. H.; Dyson, T. M.; Edwards, D. I.; Tocher, D.
A. Anti-Cancer Drug Des. 1992, 7, 3.
(17) Sheldrick, W. S.; Heeb, S. Inorg. Chim. Acta 1990, 168, 93.
(18) Morris, R. E.; Aird, R. E.; Murdoch, P. d. S.; Chen, H.; Cummings,
J.; Hughes, N. D.; Parsons, S.; Parkin, A.; Boyd, G.; Jodrell, D. I.; Sadler,
P. J. J. Med. Chem. 2001, 44, 3616.
(19) (a) Huxham, L. A.; Cheu, E. L. S.; Patrick, B. O.; James, B. R.
Inorg. Chim. Acta 2003, 352, 238. (b) Gopal, Y. N. V.; Jayaraju, D.;
Kondapi, A. K. Biochemistry 1999, 38, 4382.
(20) (a) Allardyce, C. S.; Dyson, P. J.; Ellis, D. J.; Heath, S. L. Chem.
Commun. 2001, 1396. (b) Akbayeva, D. N.; Gonsalvi, L.; Oberhauser, W.;
Peruzzini, M.; Vizza, F.; Bru¨ggeller, P.; Romerosa, A.; Sava, G.; Bergamo,
A. Chem. Commun. 2003, 264.
(21) For a recent review on pta coordination chemistry used in catalysis
and medicine, see: Phillips, A. D.; Gonsalvi, L.; Romerosa, A.; Vizza, F.;
Peruzzini, M. Coord. Chem. ReV. 2004, 248, 955.
(22) Dorcier, A.; Dyson, P. J.; Gossens, C.; Rothlisberger, U.; Scopelliti,
R.; Tavernelli, I. Organometallics 2005, 24, 2114.
(23) Scolaro, C.; Bergamo, A.; Brescacin, L.; Delfino, R.; Cocchietto,
M.; Laurenczy, G.; Geldbach, T. J.; Sava, G.; Dyson, P. J. J. Med. Chem.
2005, 48, 4161.
(η5-C5Me5){pta(H)}ptaCl2]+, presumably in which the pta ligand
is N-protonated. In fact, the N-sites of metal-pta complexes are
readily protonated in protic solvents such as alcohols and

4092 Organometallics, Vol. 25, No. 17, 2006
Dorcier et al.
Figure 2. ORTEP representation of the complex cation 4 drawn
at 30% probability ellipsoids with solvate and anion components
removed. Key bond lengths (Å) and angles (deg) for 4‚Cl: Rh1-
Cp* (centroid) 1.859, Rh1-P1 2.291(2), Rh1-P2 2.288(2), Rh1-
Cl1 2.405(2), Rh1-Cl2 6.005(2), Cp*-Rh1-P1 125.13, Cp*-
Rh1-P2 126.53, Cp*-Rh1-Cl1 125.13, P1-Rh-P2 96.93(7).
Key bond lengths (Å) and angles (deg) for 4‚BPh4: Rh1-Cp*
(centroid) 1.860, Rh1-P1 2.2970(8), Rh1-P2 2.3059(8), Rh1-
Cl1 2.4087(8), Cp*-Rh1-P1 125.42, Cp*-Rh1-P2 126.53, Cp*-
Rh1-Cl1 126.88, P1-Rh-P2 97.44(3).
Figure 1. ORTEP representations of 3 drawn at 30% probability
ellipsoids. Key bond lengths (Å) and angles (deg): Rh1-Cp*
(centroid) 1.804, Rh1-P2 2.286(1), Rh1-Cl1 2.410(1), Rh1-Cl2
2.417(1), Cp*-Rh1-P2 29.75, Cp*-Rh1-Cl1 124.84, Cp*-
Rh1-Cl2 124.53, P2-Rh-Cl1 83.95(4).
water,24 which is, apparently, extremely facile in the case of
rhodium-pta complexes. For example, both [RhCl{pta(H)}(pta)2]-
Cl and [RhCl{pta(H)}3(pta)]Cl3 are prepared by refluxing either
RhCl3 or RhCl(pta)3 with stoichiometric equivalents of pta in
EtOH and H2O, respectively.25 Compound 4 is naturally
charged, and in methanol, the only peak observed in the ESI-
MS is that of the parent cation at m/z 587. Two Rh(I) complexes
bearing η5-C5H5 and one or two pta ligands, namely, [Rh(η5-
C5Me5)(CO)(pta)] (5) and [Rh(η5-C5H5)(pta)2] (6) were also
synthesized by substituting one CO or two PPh3 ligands from
suitable Rh(I) precursors [Cp′RhL2] (Cp′ ) Cp, Cp*; L ) CO,
PPh3) (Scheme 2).
[Cp*RhCl2{P(OPh)3}] are known.27 A comparison of the former
with 3 reveals a slightly shorter Rh-P bond ([Cp*RhCl2-
{P(OEt)3}], 2.268(3) Å; 3, 2.286(1) Å). The only previously
reported Rh-P(pta) distance within the complex Rh(pta)(pta-
me)3Cl, where the metal center has a +1 oxidation state, is
significantly shorter (2.206(1) Å) than ones for the Cp*Rh(III)
species reported in this paper.25 Conversely, the Rh-Cl bond
distances in [Cp*RhCl2{P(OEt)3}] differ from one another
(2.381(3), 2.411 (4) Å), whereas in 3, they are equivalent (2.410-
(1) and 2.417(1) Å). This is perhaps due to intermolecular
interactions within the lattice system. The molecular packing
of 3 is characterized by a small number of contacts, the shortest
being between a nitrogen center of pta and a methyl hydrogen
of the η5-C5Me5 ligand, 2.547 Å, and a longer contact between
one chlorine substituent and a CH2 group of pta, 2.898 Å.
Interestingly, the other Cl group has no other interactions aside
from its bonding to the metal, which is also the case for
[Cp*RhCl2{P(OEt)3}].
The 1H and 31P{1H} NMR spectra of 3 and 4 in CDCl3 agree
with the proposed formulas. The methyl protons of the η5-C5-
Me5 group are observed at 1.69 ppm as a doublet (JHP ) 3.5
Hz) for 3 and at 1.84 ppm as a triplet (JHP ) 3.0 Hz) for 4.
Peaks of single resonances between 4.2 and 4.9 ppm correspond
to the methylene protons within the pta ligand, and the simple
chemical shift pattern indicates that the ligand is unprotonated.
The solution 31P{1H} NMR spectrum of 3 and 4 exhibits a
doublet resonance at -32.37 ppm (JPRh ) 141 Hz) and at
-28.68 ppm (JPRh ) 131 Hz), respectively.
The structure of the cation 4 is depicted in Figure 2, with
key bonding parameters listed in the caption. Two different
crystal structures have been obtained, which differ by the type
of anion and solvate contained within the lattice.
The two rhodium(I) complexes [Rh(η5-C5Me5)(CO)(pta)] and
[Rh(η5-C5H5)(pta)2] exhibit typical doublet resonances in their
solution 31P{1H} NMR spectra at -36.8 ppm (1JRhP ) 187.2
Hz) and -25.32 (d, JPRh ) 205.0 Hz). The high magnitude of
The Rh-Cl and Rh-P bond distances in 3 and the two salts
of 4 are essentially equivalent. The major difference between
the mono- and bis-pta species lies in the coordination of the
η5-C5Me5 ring to the Rh center. In 3, the two methyl substituents
that eclipse the pta ligand are lifted out of the C5 ring plane, by
5.49°. In contrast, the presence of two pta units in 4 results in
a lengthening of the Rh-Cp(centroid) distance as compared to
3 [1.861 versus 1.804 Å] and in all of the methyl groups being
bent above the C5 ring plane, ranging from 9.86° to 3.07°.
Similarly, a CCDC search on [RhCp*Cl(PR3)2]+ revealed only
1
the JRhP is in line with the presence of a Rh(I) center.26
Single crystals of 3, 4‚Cl, and 4‚BPh4 were obtained (see
Experimental Section) and their structures determined by X-ray
crystallography. The molecular structure of 3 is depicted in
Figure 1, with key bonding parameters listed in the caption.
A survey of the CCDC for compounds where a monocoor-
dinate phosphine is attached to the Cp*RhCl2 moiety reveals
in total 45 structures. Surprisingly, the search demonstrated very
few systems containing a simple homoleptic phosphine such
as PPh3 or PMe3; however, both [Cp*RhCl2{P(OEt)3}] and
four systems with nonchelating phosphines, where PR3
)
P(Ph2)CH2CH2NEt2, P(PPh)2CCPh, P(Ph)2CHCH2, and P(Ph)2-
CH2CHCH2. A comparison of 4 with the isoelectronic complex
[Ru(η5-C5Me5)(pta)2Cl]24 reveals the primary difference between
(24) Akbayeva, D. N.; Moneti, S.; Peruzzini, M.; Gonsalvi, L.; Ienco,
A.; Vizza, F. C. R. Chim. 2005, 8, 1491.
(25) Darensbourg, D. J.; Stafford, N. W.; Joo´, F.; J. H. Reibenspies, J.
Organomet. Chem. 1995, 488, 99.
(26) See for example: Bianchini, C.; Masi, D.; Meli, A.; Peruzzini, M.;
Zanobini, F. J. Am. Chem. Soc. 1988, 110, 6411.
(27) Ogo, S.; Suzuki, T.; Ozawa, Y.; Isobe, K. Inorg. Chem. 1996, 35,
6093.

Rhodium and Osmium RAPTA Analogues
Organometallics, Vol. 25, No. 17, 2006 4093
the two molecules is a 5° rotation of the cyclopentadienyl ring
with respect to the Ru-Cl bond. Interestingly, for the bis-pta
species, the Ru-P and Rh-P bond lengths are equal [Ru(η5-
C5Me5)(pta)2Cl: av 2.285(4) Å; 4BPh4-: 2.297(1), 2.288(1)
Å; 4Cl-: 2.291(2), 2.288(2) Å], but the Ru-Cl bond is longer
than that of Rh-Cl [Ru(η5-C5Me5)(pta)2Cl: av 2.467(2) Å,
4BPh4-: 2.4055(8) Å, 4Cl-: 2.405(2) Å]. Similarly, the Ru-
Cp*(centroid) distance (1.860, 1.862 Å) is equivalent to the Rh-
Cp*(centroid) separation [1.865 Å], and a similar amount of
bending of the methyl groups above the C5 plane is also
observed. In 3 the P-Rh-P bond angle (4BPh4-: 96.61(1)°,
4Cl-: 96.93(7)°) is slighter greater than the P-Ru-P bond
angle (av 93.34(9)°), but whereas the P-Rh-Cl angles (84.30-
(1)°, 85.61(1)°) are nearly equal in 4, for [Ru(η5-C5Me5)(pta)2-
Cl, the Cl is pushed to one side, resulting in unequal P-Ru-
Cl bond angles (av 84.33(10)° versus 90.82(11)°).
In Vitro Evaluation. Compounds 1-5 (see Figure 1) were
evaluated in vitro by testing for inhibition of cell proliferation
activity against the HT29 colon carcinoma, A549 lung carci-
noma, and T47D breast carcinoma cell lines. Due to air
sensitivity, it was not possible to test complex 6 along with
1-5. The effects of 1-5 on the growth of these cell lines were
evaluated after 72 h treatment, and the results from these
experiments are displayed in Figure 3. The experiments were
repeated twice for all the compounds, and the corresponding
IC50 values resulting from an average of the two experiments
are listed in Table 1 for the three cell types.
From Figure 3 and Table 1 it can be seen that for the HT29
cell line RAPTA-C, the model compound in this study, and the
osmium [Os(η6-C10H14)(pta)Cl2] and rhodium [Rh(η5-C5Me5)-
(pta)Cl2] analogues exhibit essentially the same order of
cytotoxicites [IC50 at 72 h 380-456 µM]. The cationic rhodium
complex [Rh(η5-C5Me5)(pta)2Cl]Cl (4) is considerably less
cytotoxic, and the rhodium(I) complex [Rh(η5-C5Me5)(CO)(pta)]
(5) showed no discernible cytotoxicity up to 200 µM in the
HT29 cells or the other cells used; it was not possible to evaluate
5 at greater concentrations due to its limited solubility in the
culture medium. Of all the compounds tested against the A549
cells, only 3 was cytotoxic to any significant extent. This result
clearly demonstrates the superior activity of the rhodium
analogue of RAPTA-C in this cell line. Previously, it has been
suggested that the greater lipophobicity of the cyclopentadienyl
ligand related to p-cymene and other C6-arene ligands reduces
uptake in cells and accordingly lowers their cytotoxicity.22 It
would appear that the pentamethylcyclopentadienyl ring does
not impede activity in this way, which is not unreasonable given
the presence of the methyl groups. In the T47D cells it is only
the rhodium complexes 3 and 4 that are cytotoxic, this time the
cationic bis-pta species being the most active.
In the in vitro cell proliferation assays the chloride concentra-
tion in the medium is ca. 120 mM, and at this concentration
hydrolysis of RAPTA-C is almost completely suppressed, as is
clear from Figure 4, which shows the 31P NMR spectra of
RAPTA-C in water with varying concentrations of chloride. At
high chloride concentration, i.e., that characteristic of the blood,
only RAPTA-C is present. In the absence of chloride the
hydrolysis product Ru(η6-C10H14)(pta)Cl(H2O)]+ dominates,
although it is in equilibrium with RAPTA-C. With the osmium
complex 2 at the chloride concentration in the medium hy-
drolysis is almost completely suppressed, as for RAPTA-C. In
contrast, with the rhodium complex 3 somewhat more than 50%
of the complex has been hydrolyzed. Despite the difference in
the extent of hydrolysis between 3 and the other two complexes,
and thus the charged state of the complexes, little difference is
Figure 3. Inhibition of cell growth proliferation in (top) HT29
cells, (middle) A549 cells, and (bottom) T47D cells, after 72 h of
exposure to the complexes.
observed with respect to their cytotoxicities. This similarity is
probably due to the fact that within the cell, where the chloride

4094 Organometallics, Vol. 25, No. 17, 2006
Dorcier et al.
complexes readily react with DNA model compounds,30 and
although ruthenium-chloride pta complexes have been reported
to interact with DNA,31 the actual target of the compounds
described herein may not be DNA since we have shown that
strong and specific interactions with proteins also occur.32
As far as we are aware, the only other class of metal-based
anticancer drugs to be systematically studied with respect to
identifying the optimum metal are the organometallic selective
estrogen receptor modulator compounds developed by Jaouen.
His group has studied the role of various metals including
titanium,33 rhenium,34 ruthenium,35 osmium,36 platinum,37 and
rhodium,38 with an optimum effect provided by iron in the
ferrocene derivative of tamoxifen.39 A series of different metals
have also been studied for metal-carbonyl-releasing compounds
that have pharmacological properties in suppressing organ graft
rejection, and while the most effective compound discovered
thus far is based on ruthenium, much of the evaluation has been
directed by synthetic limitations.40 This research, like that
described herein, demonstrates the need for a more thorough
and methodical approach with respect to metal selection in
medicinal organometallic chemistry, which is less well inves-
tigated than the related area of medicinal coordination com-
plexes.4
Table 1. Inhibition of Cell Proliferation as Determined by
the MTT Assay
IC50 (µM)
complex
HT29
A549
T47D
1
2
3
441
456
380
1105
1430
584
1034
>1600
512
4
>1600
>200
956
>200
346
>200
5a
a The compound was tested at a maximum concentration of 200 µM,
which is its limit of solubility in the culture medium.
Experimental Section
General Procedures. All synthetic procedures were carried out
using standard Schlenk glassware under an inert atmosphere of dry
nitrogen. The ligand pta41 and the complexes [Rh(η5-C5Me5)-
(CO)2],42 [(η5-C5Me5)RhCl(µ2-Cl)]2,43 [Rh(η5-C5H5)(PPh3)2],44 [Ru-
(η6-C10H14)(pta)Cl2],20 and [Os(η6-C10H14)(pta)Cl2]22 were prepared
as described in the literature. Other reagents were obtained from
commercial suppliers and used without further purification. Solvents
were distilled and degassed according to standard procedures.45
Figure 4. Influence of chloride concentration on the hydrolysis
of RAPTA-C determined by 31P NMR spectroscopy. [RAPTA-C]
) 5 mM, [NaNO3] ) 1 M: (a) [Cl-] ) 0 mM, (b) [Cl-] ) 5 mM,
(c) [Cl-] ) 22.7 mM, (d) [Cl-] ) 104 mM, (e) [Cl-] ) 200 mM.
The chloride in solution resulting from hydrolysis of RAPTA-C is
not included. A corresponds to unmodified RAPTA-C, B to Ru-
(η6-C10H14)(pta)Cl(H2O)]+.
(30) (a) Smith, D. P.; Baralt, E.; Morales, B.; Olmstead, M. M.; Maestre,
M. F.; Fish, R. H. J. Am. Chem. Soc. 1992, 114, 10647. (b) Smith, D. P.;
Griffin, M. T.; Olmstead, M. M.; Maestre, M. F.; Fish, R. H. Inorg. Chem.
1993, 32, 4677. (c) Smith, D. P.; Kohen, E.; Maestre, M. F.; Fish, R. H.
Inorg. Chem. 1993, 32, 4119. (d) Smith, D. P.; Olmstead, M. M.; Noll, B.
C.; Maestre, M. F.; Fish, R. H. Organometallics 1993, 12, 593.
(31) Romerosa, A.; Campos-Malpartida, T.; Lidrissi, C.; Saoud, M.;
Serrano-Ruiz, M.; Peruzzini, M.; Garrido-Ca´rdenas, J. A.; Garc´ıa-Maroto,
F. Inorg. Chem. 2006, 45, 1289.
(32) Allardyce, C. S.; Dyson, P. J.; Ellis, D. J.; Salter, P. A.; Scopelliti,
R. J. Organomet. Chem. 2003, 668, 35.
(33) Top, S.; Kaloun, E. B.; Vessie`res, A.; La¨ıos, I.; Leclercq, G.; Jaouen,
G. J. Organomet. Chem. 2002, 643-644, 350.
(34) (a) Top, S.; Vessie`res, A.; Pigeon, P.; Rager, M. N.; Huche´, M.;
Salomon, E.; Cabestaing, C.; Vaissermann, J.; Jaouen, G. ChemBioChem
2004, 5, 1104. (b) Jaouen, G.; Top, S.; Vessie`res, A.; Pigeon, P.; Leclercq,
G.; La¨ıos, I. Chem. Commun. 2001, 383.
(35) Pigeon, P.; Top, S.; Vessie`res, A.; Huche´, M.; Hillard, E. A.;
Salomon, E.; Jaouen, G. J. Med. Chem. 2005, 48, 2814.
(36) Chan K. H.; Leong, W. K.; Jaouen, G.; Leclerq, L.; Top, S.;
Vessieres, A. J. Organomet. Chem. 2006, 691, 9.
concentration is much lower, complexes 1, 2, and 3 are nearly
completely hydrolyzed and therefore equally reactive toward
the potential target, i.e., DNA or RNA.
RAPTA-C is the prototype compound that shows considerable
potential as an antimetastasis agent in vivo.23 Attempts to
improve its efficacy by modification of the arene ligand for
functionalized arenes28 and substitution of the arene by chelating
six-electron donor ligands29 have already been attempted. The
results emanating from this study, however, indicate that in
addition to ligand modification it may prove interesting to study
compounds based on different metals; certainly the osmium and
rhodium compounds described herein are worth studying further.
As mentioned in the Introduction, other types of ruthenium-
(II)-arene compounds are under investigation for their antitumor
properties. Some [Ru(η6-arene)(en)Cl] (en ) ethylenediamine)
complexes exhibit IC50 values as low as those as cisplatin in
certain types of cancer cells, whereas the osmium analogues
are not cytotoxic, apparently due to the formation of the inactive
hydroxyl species [Os(η6-arene)(en)(OH)]. Such a limitation is
not deemed likely with the bis-chloride osmium compound 2
described herein. Perhaps it should be no surprise that the
rhodium compounds described herein are active since Fish has
shown that rhodium(III)-pentamethylcyclopentadienyl aqua
(37) Top, S.; Kaloun, E. B.; Vessie`res, A.; Leclercq, G.; La¨ıos, I.;
Ourevitch, M.; Deuschel, C.; McGlinchey, M. J.; Jaouen G. ChemBioChem.
2003, 4, 754.
(38) Fish, R. H. Personal communication.
(39) (a) Top, S.; Tang, J.; Vessie`res, A.; Carrez, D.; Provot, C.; Jaouen,
G. Chem. Commun. 1996, 955. (b) Jaouen, G.; Top, S.; Vessie`res, A.;
Leclercq, G.; Quivy, J.; Jin, L.; Croisy, A. C. R. Acad. Sci. Paris Ser. IIc
2000, 3, 89.
(40) Johnson, T. R.; Mann, B. E.; Clark, J. E.; Foresti, R.; Green, C.;
Motterlini R. Angew. Chem., Int. Ed. 2003, 42, 2.
(41) (a) Daigle, D. J.; Pepperman, A. B., Jr.; Boudreaux G. J. Heterocycl.
Chem. 1974, 11, 1085. (b) Daigle, D. J. Inorg. Synth. 1998, 32, 40.
(42) Maitlis, P. M.; Kang, J. W. J. Organomet. Chem. 1971, 26, 393.
(43) Kang, J. W.; Moseley, K.; Maitlis, P. M. J. Am. Chem. Soc. 1969,
91, 5970.
(44) Yamaraki, H.; Hagihara, N. Bull. Chem. Soc. Jpn 1971, 44, 2260.
(45) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; 1988.
(28) Scolaro, C.; Geldbach, T. J.; Rochat, S.; Dorcier, A.; Gossens, C.;
Bergamo, A.; Cocchietto, M.; Tavernelli, I.; Sava, G.; Rothlisberger, U.;
Dyson, P. J. Organometallics 2006, 25, 756.
(29) Serli, B.; Zangrando, E.; Gianferrara, T.; Scolaro, C.; Dyson, P. J.;
Bergamo, A.; Alessio, E. Eur. J. Inorg. Chem. 2005, 3423.

Rhodium and Osmium RAPTA Analogues
Organometallics, Vol. 25, No. 17, 2006 4095
Table 2. Crystal Data and Details of the Structure Determination for 3, 4‚Cl, and 4‚BPh4
3
4‚Cl
4‚BPh4
chemical formula
fw
cryst syst
space group
a (Å)
C16H27Cl2N3PRh
466.19
orthorhombic
P212121
8.500(3)
3.382(6)
16.361(9)
90
C22H40Cl2N6O2P2Rh
656.35
triclinic
P1h
9.1729(18)
12.325(3)
13.028(3)
83.11(3)
1362.5(6)
2
C46.70H61.81BClN6O0.705P2Rh
929.54
monoclinic
P21/c
11.9850(10)
30.388(2)
13.0020(9)
107.800(7)
4508.6(6)
4
b (Å)
c (Å)
â (deg)
volume (Å3)
Z
1861.0(15)
4
1.664
D
calc (g cm-3
)
1.607
1.369
F(000)
952
1.293
200
684
0.973
150
1946
0.551
200
µ (mm-1
)
temp (K)
wavelength (Å)
no. of measd reflns
no. of unique reflns
no. of unique reflns [I > 2σ(I)]
no. of data/params
Ra [I > 2σ(I)]
0.71073
2310
2255
0.71073
14 259
4815
0.71073
101 216
8037
7960
8037/530
0.0454
2180
2640
2255/213
0.0190
0.0480
1.00
4815/323
0.0635
0.1218
0.982
wR2a (all data)
0.0971
1.164
GoFb
2
2
2
2
2
a R ) ∑||Fo| - |Fc||/∑|Fo|, wR2 ) {∑[w(Fo - Fc )2]/∑[w(Fo )2]}1/2
.
b GoF ) {∑[w(Fo - Fc )2]/(n - p)}1/2 where n is the number of data and p is the
number of parameters refined.
Infrared spectra were recorded on a Perkin-Elmer Spectrum BX
series FT-IR spectrometer in KBr disks. The 1H and 13C{1H} NMR
spectra were recorded on a Bruker AC200 spectrometer operating
at 200.13 and 50.32 MHz, respectively. Peak positions are relative
to tetramethylsilane and were calibrated against the residual solvent
resonance (1H) or the deuterated solvent multiplet (13C). 31P{1H}
NMR spectra were recorded on the same instrument operating at
81.01 mHz. Chemical shifts were measured relative to external 85%
H3PO4, with downfield shifts considered positive. All the NMR
spectra were recorded at room temperature (20 °C) unless otherwise
stated. Elemental analyses (C, H) were performed using a Carlo
Erba model 1106 elemental analyzer by the Microanalytical Service
of the Department of Chemistry at the University of Florence. ESI-
MS of the complexes were obtained on a Thermofinigan LCQ Deca
XP Plus quadrupole ion trap instrument set in positive mode
(solvent: methanol; flow rate: 5 µl/mn; spray voltage: 5 kV;
capillary temperature: 100 °C; capillary voltage: 20 V), as
described previously.46 The percentage given in brackets belongs
to the relative intensity of the peaks.
Synthesis of [Rh(η5-C5Me5)(pta)Cl2], 3. A solution of [(η5-C5-
Me5)RhCl(µ2-Cl)]2 (100 mg, 0.162 mmol) and pta (51 mg, 0.324
mmol) in CHCl3 (25 mL) was left standing for 12 h. After
evaporation, the residue was washed with ether (3 × 10 mL) and
CHCl3 (3 × 5 mL) to give a fine red microcrystalline powder (110
mg, 0.236 mmol, 73%). Recrystallization from methanol gave red
crystals of 3 suitable for X-ray diffarction analysis.
1H NMR (CDCl3): δ 1.69, (d, Cp-CH3, JHP ) 3.5 Hz), δ 4.33
(s, PCH2N), δ 4.51 (s, NCH2N). 31P{1H} NMR (CDCl3): δ -32.37
(d, JPRh ) 141 Hz). ESI-MS: m/z ) 430.0 [Rh(η5-C5Me5)(pta)-
(Cl)]+ (10%), m/z ) 465.9 [Rh(η5-C5Me5){pta(H)}ptaCl2]+ (100%),
and m/z ) 932.6 [Rh2(η5-C5Me5)2{pta(H)}pta(pta)Cl4]+ (38%).
Anal. Calcd for C16H27N3Cl2PRh: C, 41.2; H, 5.8; N, 9.0. Found:
C, 41.0; H, 5.9; N, 9.1.
-28.68 (d, JPRh ) 131 Hz). ESI-MS m/z ) 587.00 [Rh(η5-C5-
Me5)(pta)2Cl]+ (100%). Anal. Calcd for C22H39N6Cl2P2Rh: C, 42.4;
H, 6.3; N, 13.5. Found: C, 42.2; H, 6.3; N, 13.3.
Synthesis of [Rh(η5-C5Me5)(pta)2Cl]BPh4, 4‚BPh4. Metathesis
reaction of 4‚Cl (0.5 mmol) with NaBPh4 (0.65 mmol) was carried
out dissolving the rhodium precursor in warm EtOH and adding
the salt in an acetone solution at room temperature. By slow
evaporation, the product was obtained as a microcrystalline powder,
which was recrystallized from EtOH/acetone to yield single dark
orange crystals suitable for X-ray structure determination. Yield:
90%. Anal. Calcd for C46H59N6BClP2Rh: C, 60.9; H, 6.6; N, 9.3.
Found: C, 61.0; H, 6.5; N, 9.1.
Synthesis of [Rh(η5-C5Me5)(CO)(pta)], 5. Solid [Rh(η5-C5Me5)-
(CO)2] (147 mg, 0.5 mmol) was added to an EtOH solution (20
mL) of pta (78.5 mg, 0.5 mmol) under nitrogen. The resulting
orange solution was stirred and refluxed for 24 h, cooled to RT,
and filtered under N2 through a cotton plug. Solvent evaporation
under a nitrogen stream gave orange crystals. These were then
filtered off, washed with petroleum ether, and dried under vacuum
(127 mg, 0.432 mmol, 60%).
1
IR (cm-1): ν (CO) 1907 (s). H NMR (CD2Cl2): δ 2.01 (dd,
Cp-CH3, JHP ) 1.9 Hz, JHRh ) 0.5 Hz, 15H), 3.98 (m, NCH2N,
6H), 4.46 (m, PCH2N, 6H). 31P{1H} NMR (CD2Cl2): δ -36.75
(d, JPRh ) 187.2 Hz). 13C{1H} NMR (CD2Cl2): δ 11.9 (s, Cp-
3
CH3), 56.37 (d, NCH2P, JCP ) 16.4 Hz), 74.0 (d, NCH2N, JCP
)
2
6.7 Hz), 99.2 ppm (dd, Cp, JCP ) 3.7 Hz, JCRh ) 1.8 Hz), 197.9
ppm (dd, CO, JCP ) 85.8 Hz, JCRh ) 25.5 Hz). Anal. Calcd for
2
C17H27N3OPRh: C, 48.2; H. 6.4; N, 9.9. Found: C, 48.2; H, 6.4;
N, 9.8.
Synthesis of [Rh(η5-C5H5)(pta)2], 6. [Rh(η5-C5H5)(PPh3)2] (180
mg, 0.26 mmol) was added to a degassed EtOH solution (150 mL)
of pta (81.7 mg, 0.52 mmol). The resulting orange solution was
stirred and refluxed under nitrogen for 2 h. The solution was then
concentrated to 8 mL, and reddish-orange crystals were slowly
formed over 12 h. These crystals were filtered off under N2, washed
with EtOH/n-hexane (1:3), and dried under vacuum.
Yield: 43.1 mg (30%). 1H NMR (CD2Cl2): δ 3.83 (br s, NCH2N,
12H), 4.45 (br s, NCH2P, 12H), 5.13 (s, Cp, 5H). 31P{1H} NMR
(CD2Cl2): δ -25.32 (d, JPRh ) 205.0 Hz). Anal. Calcd for
C17H29N6P2Rh: C, 42.3; H, 6.1; N, 17.4. Found: C, 42.8; H, 6.7;
N, 17.0.
Synthesis of [Rh(η5-C5Me5)(pta)2Cl]Cl, 4‚Cl. A mixture of [(η5-
C5Me5)RhCl(µ2-Cl)]2 (100 mg, 0.162 mmol) and pta (110 mg, 0.700
mmol) in CHCl3 (25 mL) was left standing for 24 h at room
temperature to give dark orange crystals, which were separated from
the mother liquor by decantation, collected by filtration in the air,
and washed with CHCl3 (2 × 3 mL) and n-pentane (2 × 5 mL)
before being dried at the pump (155 mg, 0.249 mmol, 77%).
1H NMR (CDCl3): δ 1.84, (t, Cp-CH3, JHP ) 3.0 Hz), δ 4.4-
4.9 (PCH2N, NCH2N, 7 peaks, 24 H). 31P{1H} NMR (CDCl3): δ
(46) Dyson, P. J.; McIndoe, J. S. Inorg. Chim. Acta 2003, 354, 68.

4096 Organometallics, Vol. 25, No. 17, 2006
Dorcier et al.
Crystallography. The data were solved using direct methods
with SHELXS and refined using SHELXL97.47 The graphics
interface package used was PLATON,48 and the figures were
generated using the ORTEP 3.0749 generation package. For both
compounds all non-hydrogen atoms have been refined anisotropi-
cally, and all the hydrogen atoms were placed using a riding model.
For compound 4‚Cl there is a disordered, partially occupied
dichloromethane solvent molecule.
Cell Culture. Human T47D breast carcinoma, A549 lung
carcinoma, and HT-29 colon carcinoma cell lines were obtained
from the American Type Culture Collection (ATCC, Manassas,
VA). All other cell culture reagents were obtained from Gibco-
BRL, Basel, Switzerland. The cells were routinely grown in DMEM
medium containing 4.5 g/L glucose, 10% fetal calf serum (FCS),
and antibiotics at 37 °C and 6% CO2. For the MTT tests, the cells
were seeded in 48-well plates (Costar, Integra Biosciences,
Cambridge, MA) as monolayers for 24-48 h in complete medium
to reach confluence, then fresh complete medium with 5% FCS
was added together with the organometallic drugs, and culture was
continued for 72 h. The test (see below) was performed for the
last 2 h without changing the culture medium.
Determination of Cell Viability. The compounds were dissolved
directly in culture medium, supplemented with 5% FCS. For 5,
the maximum concentration test was 200 µM due to the limited
solubility of the compound in culture medium. Cell viability was
determined using the MTT assay, which allows the quantification
of the mitochondrial activity in metabolically active cells. Following
drug exposure, MTT (final concentration 0.2 mg/mL) was added
to the cells for 2 h, then the culture medium was aspirated and the
violet formazan precipitate dissolved in 0.1 N HCl in 2-propanol.
The optical density, which is directly proportional to number of
surviving cells, was quantified at 540 nm using a multiwell plate
reader (iEMS Reader MF, Labsystems), and the fraction of
surviving cells, expressed as an average of three independent tests,
was calculated from the absorbance of untreated control cells.
Acknowledgment. We thank the Swiss National Science
Foundation, the Ligue Suisse Contre le Cancer (Swiss Cancer
League), the EPFL, and the EC through the contract MRTN-
CT-2003-503864 (AQUACHEM) for financial support. This
work was also supported by COST D29 (WG009/03).
Supporting Information Available: Crystallographic informa-
tion files (CIF) for 3, 4‚Cl, and 4‚BPh4. This material is available
free of charge via the Internet at http://pubs.acs.org.
(47) Sheldrick, G. M. SHELXS97 and SHELXL97; University of Go¨t-
tingen: Germany, 1997.
(48) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7.
(49) Farrugia, L. J. ORTEPIII for Windows. J. Appl. Crystallogr. 1997,
30, 565.
OM060394O