Can metal complexes serve as hypoxia activated prodrugs? Investigations of a Co(III) complex of the MMP inhibitor marimastat

For many years proof that the hypoxic nature of malignant tumours can be used to selectively target anticancer drugs has been sought. Several classes of potential redox activated anticancer drugs have been developed to take advantage of the reducing environment resulting from the hypoxia. Drug complexes with redox active metal centres as carriers have been investigated, but have largely been employed with cytotoxic drugs that require release of the drug intracellularly, complicating the design of such complexes. MMP inhibitors, a new class of anticancer drug, conversely act in the extracellular environment and we have investigated inhibitor complexes with several redox active transition metals. Marimastat is an MMP inhibitor with potent in-vitro antimetastatic activity and was recently in Phase III clinical trials for a variety of cancer types. We have synthesised a Co(II1) complex of marimastat incorporating the tetradentate ligand tpa (tris(2-methylpyridyl)amine) as a carrier ligand. The complex was structurally characterised in the solid state by single crystal X-ray diffraction, the first example of a crystal structure containing marimastat. 2D COSY and NOESY NMR spectra showed that the complex exists in two isomeric forms in solution, corresponding to the cis and trans isomers yet only crystallises in one of these forms. Biological testing of the complex in mice with 4T1.2 tumours showed interesting and unexpected outcomes. Initial results of the tumour growth inhibition study showed that a significant inhibition of growth was exhibited by the complex over the free inhibitor and the control. However, the metastatic potential of both free marimastat and the complex were higher than the control indicating likely problems with the experimental protocol. Further experiments are needed to determine the potential of such complexes as hypoxia activated prodrugs but there appears at least to be some promise.

Cyanocobalt(III) complexes of penta- and tetradentate-coordinated macrocyclic hexaamines

The pendent-arm macrocyclic hexaamine trans-6,13-dimethyl-1,4,8,11-tetraazacyclotetradecane-6,13-diamine (L) may coordinate in tetra-, penta- or hexadentate modes, depending on the metal ion and the synthetic procedure. We report here the crystal structures of two pseudo-octahedral cobalt(III) complexes of L, namely sodium trans-cyano(trans-6,13-dimethyl-1,4,8,11-tetraazacyclotetradecane-6,13-diamine)cobalt(III) triperchlorate, Na[Co(CN)(C13H30N6)](ClO4)(3) or Na{trans-[CoL(CN)]}(ClO4)(3), (I), where L is coordinated as a pentadentate ligand, and trans-dicyano(trans-6,13-dimethyl-1,4,8,11-tetraazacyclotetradecane-6,13-diamine) cobalt (III) trans-dicyano (trans-6,13-dimethyl-1,4,8,11-tetraazacyclotetradecane-6,13-diaminium)cobalt(III) tetraperchlorate tetrahydrate, [Co(CN)(2)(Cl4H32N6)][Co(CN)(2)(Cl4H30N6)](ClO4)(4)•-4H(2)O or trans-[CoL(CN)(2)]trans-[Co(H2L)(CN)(2)] (ClO4)(4)•-4H(2)O, (II), where the ligand binds in a tetradentate mode, with the remaining coordination sites being filled by C-bound cyano ligands. In (I), the secondary amine Co-N bond lengths lie within the range 1.944 (3)-1.969 (3) &ANGS;, while the trans influence of the cyano ligand lengthens the Co-N bond length of the coordinated primary amine [Co-N = 1.986 (3) &ANGS;]. The Co-CN bond length is 1.899 (3) &ANGS;. The complex cations in (11) are each located on centres of symmetry. The Co-N bond lengths in both cations are somewhat longer than in (I) and span a narrow range [1.972 (3)-1.982 (3) &ANGS;]. The two independent Co-CN bond lengths are similar [1.918 (4) and 1.926 (4) &ANGS;] but significantly longer than in the structure of (1), again a consequence of the trans influence of each cyano ligand.