Reduction of in vivo lung metastases by dinuclear ruthenium complexes is coupled to inhibition of in vitro tumour invasion.

Mononuclear ruthenium-dmso compounds showed interesting antimetastatic properties on experimental models of solid tumours. In line with the interesting results with multinuclear platinum complexes, which proved to overcome cisplatin resistance, we thought it worthwhile to test the pharmacological properties of some dinuclear ruthenium complexes to ascertain the possible advantages due to the introduction of a second metal centre over NAMI-A and its mononuclear analogues. These compounds belong to the general formula X2[[RuCl4(dmso-S)]2(mu-L)] or [X][[RuCl4(dmso-S)](mu-L)[RuCl3(dmso-S)(dmso-O)]] where L is a nitrogen donor ligand (pyrazine; pyrimidine; 4,4'-bipyridine; 1,2-bis(4-pyridyl)ethane; 1,2-bis(4-pyridyl) ethylene; 1,3-bis(4-pyridyl)propane) and X a counterion. We focused on parameters related to metastatic ability such as gelatinase activity, detected by zymography, and invasive potential, measured by means of a transwell chamber. These activities were correlated to the ability to inhibit tumour metastases in vivo. All dinuclear complexes, except compound D8 ([NH4]2[[RuCl4(dmso-S)]2(mu-pyz]), decrease the number of tumour cells that cross a matrigel barrier, and inhibit MMP-9 gelatinolytic activity at concentrations lower than that of NAMI-A and of other mononuclear ruthenium complexes. In vivo compounds D5 (Na2[[RuCl4(dmso-S)]2(mu-ethylbipy)]) and D7 ([NH4][[RuCl4(dmso-S)](mu-pyz)[RuCl3(dmso-S) (dmso-O)]]) show anti-metastasis activity, at two dose levels, with mild or null effect on primary tumour growth; compound D8 is the weakest active. All compounds tend to accumulate in liver and kidneys, rather than in tumour and lungs. However, compound D5, the most active in vitro on invasion and gelatinases and active in vivo on metastasis, is better concentrated in the lungs than compound D8 which is less active or inactive in vitro and in vivo. Histological analysis show liver, as well as kidney toxicities that limit in vivo activity. These data thus suggest dinuclear ruthenium complexes as promising anti-invasive agents for cancer treatment.

Ruthenium-based NAMI-A type complexes with in vivo selective metastasis reduction and in vitro invasion inhibition unrelated to cell cytotoxicity.

A series of analogues of NAMI-A, a reference compound active on solid tumor metastases, were synthesized (NAMI-A type complexes). They share the same chemical structure of NAMI-A, and differ from it in the nature of the coordinated nitrogen ligand, such as pyrazole, thiazole and pyrazine, which are less basic than imidazole. This modification confers to the new NAMI-A type complexes a better stability in aqueous solution compared to the parent compound, a very important characteristic for a class of compounds that, with NAMI-A, is currently completing a phase I clinical trial at the Netherlands Cancer Institute of Amsterdam. Cytotoxicity and the effects on cell cycle and invasion were investigated on TS/A, B16-F10 and MCF-7 tumor cell lines, while the inhibition of lung metastases was determined on the mouse experimental tumors Lewis lung carcinoma and MCa mammary carcinoma. The new complexes show a pharmacological activity very similar to that of the parental compound NAMI-A: in vitro they are devoid of meaningful cytotoxicity against tumor cells, and in vivo they inhibit metastasis formation and growth approximately to the same extent as NAMI-A. Thus the new NAMI-A type complexes retain the same potent characteristic of NAMI-A to selectively interact with solid tumor metastases. However, compared to NAMI-A they do not stop cell cycle progression at G2-M level and are more active in preventing the spontaneous invasion of Matrigel by tumor cells exposed for 1 h to 10(-4) M concentration. Globally, these complexes take advantage of the knowledge on NAMI-A and appear particularly interesting for future clinical handling and applications.

Solution, solid state and biological characterization of ruthenium(III)-DMSO complexes with purine base derivatives.

Two new complexes of Ru(III) with purine base derivatives, [mer-RuCl(3)(acv)(DMSO-S)(C(2)H(5)OH)].C(2)H(5)OH (1) (acv=acyclovir, DMSO=dimethyl sulfoxide) and [trans-RuCl(4)(guaH)(DMSO-S)].2H(2)O (2) (guaH=protonated molecule of guanine), were prepared from the same Ru(III) precursor, [trans-RuCl(4)(DMSO-S)(2)](-), by substitution of one DMSO-S. Coordination of acv induced also replacement of one chloride by an ethanol molecule. This reactivity difference was explained by striking contrasts in the hydrogen bonding schemes of the two complexes, evidenced in their X-ray crystal structures. In 1 the guanine derivative acyclovir is coordinated to ruthenium through the N(7) atom, while in 2 the protonated guanine molecule is bound through the N(9) atom. Both complexes were also characterized by various physico-chemical methods in the solid state and in the solution. In vitro, the biological activity of 2 and of the previously described complexes [mer-RuCl(3)(acv)(DMSO-S)(CH(3)OH)].0.5CH(3)OH (3) and [mer-RuCl(3)(acv)(DMSO-S)(H(2)O)].H(2)O (4) on tumour cells appear to be very similar to that of NAMI-A (NAMI-A=[ImH][trans-RuCl(4)(DMSO-S)Im]). All compounds are only weakly active on tumour cell proliferation but show an interesting proadhesive effect that suggest possible activity on tumour malignancy.

Replacement of chlorides with dicarboxylate ligands in anticancer active Ru(II)-DMSO compounds: a new strategy that might lead to improved activity.

A series of new Ru(II)-DMSO complexes containing dicarboxylate ligands (dicarb), namely, oxalate (ox), malonate (mal), methylmalonate (mmal), dimethylmalonate (dmmal), and succinate (suc), have been synthesized and structurally characterized. These compounds were prepared from the known Ru(II)-Cl-DMSO anticancer complexes cis,fac-[RuCl2(DMSO-S)3(DMSO-O)] (1) and trans-[RuCl2(DMSO-S)4] (2) and from the chloride-free precursor fac-[Ru(DMSO-S)3(DMSO-O)3][CF3SO3]2 (3), with the aim of assessing how the nature of the anionic ligands influences the biological activity of these species. Basically, the investigated ligands can be divided into two groups. The reaction of either 1 or 2 with K2(dicarb) (dicarb = ox, mal, mmal) yielded preferentially the mononuclear species [K]fac-[RuCl(DMSO-S)3(eta2-dicarb)] (dicarb = mal, 6; mmal, 9; ox, 14) that contains a chelating dicarboxylate unit and a residual chloride. Likewise, when 3 was used as a precursor, the neutral mononuclear species fac-[Ru(DMSO-O)(DMSO-S)3(eta2-dicarb)] (dicarb = mal, 7; mmal, 10; ox, 16), which contains a DMSO-O ligand in the place of Cl-, was obtained. On the contrary, K2(suc) and K2(dmmal) yielded preferentially the dinuclear species [fac-Ru(DMSO-S)3(H2O)(mu-dicarb)]2 (dicarb = dmmal, 11; suc, 13), with two bridging dicarboxylate moieties. The two water molecules in anti geometry have strong intramolecular H-bonding with the non-coordinated oxygen atoms of the carboxylate groups. The solid-state X-ray structural data showed that the preferential binding mode of the investigated dicarboxylates, either bridging (mu) or chelating (eta2), is dictated mainly by steric reasons. Oxalate, unlike the other dicarboxylates, has also the bridging bis-chelate (eta4,mu) coordination mode available: this was found in the dinuclear species [{fac-RuCl(DMSO-S)3}2(eta4,mu-ox)] (15) and [{fac-Ru(DMSO-O)(DMSO-S)3}2(eta4,mu-ox)][CF3SO3]2 (17). We also isolated the unprecedented neutral metallacycle, [fac-Ru(DMSO-S)3(eta3,mu-ox)]4 (18), in which each oxalate unit has one unbound oxygen atom. The new complexes were thoroughly characterized by 1-D (1H and 13C) and 2-D (H-H- COSY and HMQC) NMR spectroscopy in solution and by IR spectroscopy in the solid state. The molecular structures of 10 compounds, 6-11, 13, 15, 17, and 18, were determined by X-ray crystallography. The behavior of selected complexes in aqueous solution was investigated by 1H NMR spectroscopy.

The unprecedented bridging coordination mode of 1,1-cyclobutane dicarboxylate (mu-cbdc-O,O') stabilized by intramolecular hydrogen bonds in ruthenium(II) complexes.

The 1,1-cyclobutane dicarboxylate ligand (cbdc), that normally binds to metal centres as a chelate (eta(2)-cbdc-O,O'), prefers to bind in an unprecedented bridging fashion (micro-cbdc-O,O') on cationic Ru(ii) centres bearing ancillary ligands (e.g. H(2)O, NH(3)) capable of making intramolecular H-bonds with the non-coordinated oxygen atoms of the carboxylate groups. Thus, the thermodynamic product of the reaction between cis,fac-[RuCl(2)(dmso-S)(3)(dmso-O)]() and cbdc in a number of different reaction conditions is the dinuclear species with two bridging cbdc units fac-[Ru(micro-cbdc-O,O')(dmso-S)(3)(H(2)O)](2) (2). Similarly, reaction of cis,fac-[RuCl(2)(dmso-S)(3)(NH(3))] (3) with cbdc yielded the corresponding dinuclear species fac-[Ru(micro-cbdc-O,O')(dmso-S)(3)(NH(3))](2) (4), in which ammonia occupies the position of the water molecule in 2. Both dinuclear species 2 and 4 were characterized by X-ray crystallography and have an anti geometry with respect to the H(2)O or NH(3) ligands. The results from the X-ray studies are consistent with the NMR spectroscopic data, indicating that the dinuclear structures observed in the solid state are maintained in solution. The mononuclear anionic complex with a chelating cbdc unit, K{fac-[RuCl(eta(2)-cbdc-O,O')(dmso-S)(3)]}(5), was isolated under appropriate conditions form the reaction of 1 with K(2)(cbdc) and was demonstrated to be an intermediate in the formation of 2.

Synthesis and structural, spectroscopic, and electrochemical characterization of new ruthenium dimethyl sulfoxide nitrosyls.

The reactivity of ruthenium(II)- and ruthenium(III)-chloride-dimethyl sulfoxide precursors and of the antimetastatic drug [ImH][trans-RuCl(4)(dmso-S)(Im)] (NAMI-A, Im = imidazole, dmso = dimethyl sulfoxide) toward NO was investigated. Treatment of [(dmso)(2)H][trans-RuCl(4)(dmso-S)(2)] and mer-RuCl(3)(dmso)(3) with gaseous NO yielded [(dmso)(2)H][trans-RuCl(4)(dmso-O)(NO)] (1) and mer,cis-RuCl(3)(dmso-O)(2)(NO) (2), respectively. Thus, coordination of the strong pi-acceptor NO induces a S to O linkage isomerization of the dmso trans to it to avoid competition for pi-electrons. In light-protected nitromethane solutions, complex 2 equilibrates slowly with the two isomers mer-RuCl(3)(dmso-S)(dmso-O)(NO) (3), with NO trans to Cl, and mer-RuCl(3)(dmso-S)(dmso-O)(NO) (4), with NO trans to dmso-O; the equilibrium mixture consists of ca. 64% 2, 3% 3, and 33% 4. Treatment of the Ru(II) precursor trans-RuCl(2)(dmso-S)(4) with gaseous NO in CH(2)Cl(2) solution yielded the nitrosyl-nitro derivative trans,cis,cis-RuCl(2)(dmso-O)(2)(NO)(NO(2)) (5). Finally, [(Im)(2)H][trans-RuCl(4)(Im)(NO)] (6) was prepared by treatment of [ImH][trans-RuCl(4)(dmso-O)(NO)] (1Im) with an excess of imidazole in refluxing acetone. The spectroscopic features are consistent with the [Ru(NO)](6) formulation for all complexes, that is, a diamagnetic Ru(II) nucleus bound to NO(+). Compounds 1, 2, 5, and 6 were characterized also by X-ray crystallography; they all show a linear nitrosyl group, with short Ru-NO bond distances consistent with a strong d(pi) --> pi NO back-bonding. An unusual inertness of O-bonded dmso was observed in compound 1. Complexes 1, 2, 3, 5, and 6 are all redox active in DMF solutions showing irreversible reductions whose peak potentials depend on the other ligands attached to the Ru metal center. The site of reduction is the NO(+) moiety. The reduced complexes are not stable and release a Cl(-) or NO(2)(-) ligand followed by the NO(*) radical. The chemical reactions following electron transfer are all fast (rate constant >100 s(-1) at 293 K). The Ru product species are not redox active within the DMF window.

Distinct effects of dinuclear ruthenium(III) complexes on cell proliferation and on cell cycle regulation in human and murine tumor cell lines.

We have examined the biological and antitumor activity of a series of dinuclear ruthenium complexes. The aim of this study was to compare the in vitro effects of these new compounds on cell proliferation, cell distribution among cell cycle phases, and the expression of some proteins involved in cell cycle regulation. Results obtained show a mild cytotoxic activity against human and murine cell lines, more evident after prolonged exposure of cell challenge. Two of the eight dinuclear complexes [namely, compounds D3 (Na(2)[(RuCl(4)(dmso-S))(2)(mu-bipy)]) and D7 ([NH(4)][(RuCl(4)(dmso-S))(mu-pyz)(RuCl(3)(dmso-S)(dmso-O))]) modify cell cycle distribution similarly to imidazolium trans-imidazoledimethylsulfoxidetetrachlororuthenate (NAMI-A), whereas the others have a low or negligible effect on this parameter. If we correlate the induction of cell cycle modifications with ruthenium uptake by tumor cells and with the modulation of proteins regulating cell cycle, we may stress that the induction of G(2)-M cell cycle arrest is related to the achievement of a threshold concentration of ruthenium inside the cells, which is dependent on the cell line being used, and that only cyclin B, among cell cycle regulating proteins examined by immunoblotting assays, appears to be significantly modified. This in vitro study shows that dinuclear ruthenium complexes may have a behavior similar to that of the monomer NAMI-A. These results encourage the future experimentation of their pharmacological properties in in vivo models.