An Unusual Dimeric Structure of a Cu(I) Bis(thiosemicarbazone) Complex: Implications for the Mechanism of Hypoxic Selectivity of the Cu(II) Derivatives Andrew R. Cowley, † Jonathan R. Dilworth,* ,† Paul S. Donnelly, † Elena Labisbal, † and Antonio Sousa ‡ Inorganic Chemistry Laboratory, UniVersity of Oxford, South Parks Road, Oxford, OX1 3QR, UK, and Departmento de Quimica Inorganica, UniVersidad de Santiago de Compostella, 15706 Santiago de Compostella, Spain Received December 6, 2001 Transition metal complexes of bis(thiosemicarbazone) (btsc) ligands (Figure 1, which also shows standard abbreviations) have been studied for nearly 50 years. 1-3 Subsequent interest has focused on the redox properties, structures and biological activity of such complexes. 4 In particular bis(thiosemicarbazone) complexes of copper(II) have been know for some time to be anti-tumor agents. 5,6 However, it is the hypoxic selectivity of certain copper bis- (thiosemicarbazones) and their use as vehicles for the delivery of radioactive copper isotopes to tumors 7,8 or leucocytes 9 that has attracted much recent attention 10 through the work of Welch and Fujibayashi and co-workers. This is exemplified by a very recent report that [ 64 Cu(II)(ATSM)] significantly improves the survival times of animals bearing human GW38 colon cancer tumors. 11 The general topic of copper-based radiopharmaceuticals has also been reviewed relatively recently. 12 The hypoxic selectivity is strongly dependent on the substituents on the carbon backbone ([Cu(ATSM)] shows good hypoxic selectivity whereas [Cu(GTS)] exhibits none), and we have studied the chemistry of these systems in some detail in an effort to understand the mechanism of hypoxic selectivity more fully. The mechanism of hypoxic selectivity of [Cu(btsc)] complexes has been discussed in terms of the redox potentials for reduction of the Cu(II) complexes to Cu(I), 13 the most selective complexes being those that are most difficult to reduce. The redox potentials are markedly dependent on the backbone substituents, and it was suggested that this variation accounted for the range of hypoxic selectivity observed. Trapping of the complexes within the cells was assumed to occur by virtue of the formation of the charged anion. The reported redox potentials were measured in dry DMF, and under these conditions two completely Nernstian reversible processes are observed, one corresponding to reduction to Cu(I) and that at positive potentials to oxidation to Cu(III). However, hypoxic cells are mildly acidic, 14 and Cu(II) btsc complexes are known to protonate. A [Cu(II)(btsc)] complex has been reported to have pK a values of 2.75 and ca. 0.8. 15 Reduction of the Cu(II) complex will further enhance the basicity of the coordinated btsc ligand. We have observed that the CVs of the Cu(II) complexes in the presence of aqueous acid are dramatically different from those in anhydrous DMF, and coupled protonation and reduction clearly cannot be neglected in the medium likely to be found within hypoxic cells. This suggests strongly that a Cu(I) anionic complex is unlikely to be formed in the reduction of the [Cu(II)(btsc)] complexes in the mildly acidic aqueous environment of hypoxic cells. The structure of [Cu(II)(ATSM)] has been determined for the first time and is shown in Figure 2. The complex comprises square planar units which are loosely associated into dimers by long Cu-S interactions. The long Cu-S interactions have been reported in the very few earlier structure determinations for Cu(btsc) complexes. 16 However there is an interesting difference for the structure of the d 10 Zn complex of ATSM which is unequivocally dimeric with five-coordinate square pyramidal Zn. 17 This suggested that the corresponding hypothetical Cu(I) anion, postulated as the species responsible for the selective trapping of Cu in hypoxic cells, might have the same structure. However, it now appears that at the concentrations used for synthesis that it is probable that this species if formed, protonates rapidly and rearranges to the dimeric species (1). We have attempted to isolate the Cu(I) species by reaction of the btsc ligands with a Cu(I) precursor. Reaction of [Cu(MeCN) 4 ]- [PF 6 ] with ATSMH 2 in MeCN unexpectedly yielded the novel dimeric species [Cu 2 (ATSMH 2 ) 2 ] 2+ , isolated as the [PF 6 ] - salt (1). All attempts to isolate an anionic species by the addition of strong base failed, suggesting that it may not in fact be stable even in aprotic media. It also proved impossible to isolate any Cu(I) species with GTS. The X-ray crystal structure of (1) (Figure 3) revealed a dimeric structure with each of the btsc ligands acting as a bidentate N-S donor to each Cu(I) ion to generate a novel helical structure which is unprecedented for bis(thiosemicarbazone) complexes. The Cu-Cu distance of 3.561 Å suggests little interaction between the two metal ions. The two components of the dimer are related by a crystallographic two-fold axis that bisects the C-C bond. Each of the ligands is twisted substantially at the C-C bond (torsion angles: N(2)-C(2)-C(2)′-N(2)′ ) 51.1°, N(4)-C(6)-C(6)′- * To whom correspondence should be addressed. E-mail: jon.dilworth@ chem.ox.ac.uk. † University of Oxford. ‡ Universidad de Santiago de Compostella. Figure 1. Structures and abbreviations for Cu(II) bis(thiosemicarbazones). Figure 2. Crystal structure of [Cu(II)(ATSM)]. Published on Web 04/18/2002 5270 9 J. AM. CHEM. SOC. 2002, 124, 5270-5271 10.1021/ja012668z CCC: $22.00 © 2002 American Chemical Society