RESUMO
Cobalt complexes have been demonstrated to target zinc fingers, as shown by investigations of Doxovir, the trade name of the [CoIII(acacen)(2-Me-Imz)2]+ drug in clinical trials. Mechanistic studies indicate zinc finger disruption by metal coordination to His residues. Other than Doxovir, a few studies have investigated other ligands and geometries for cobalt complexes for zinc finger targeting. Tripod ligands demonstrated good zinc and cobalt chelation. In this manuscript, we report the ability of CoII and CoIII complexes of tri(2-pyridylmethyl)amine and N,N-di(2-pyridylmethyl)glycinate to disrupt zinc fingers. The results obtained by mass spectrometry and X-ray absorption spectroscopy demonstrate that the complexes were able to remove zinc from the zinc fingers. The product was oxidised apo-peptide. In contrast, the ligands themselves were able to remove zinc, and they did not promote oxidation, resulting in free Cys residues. Cobalt finger adducts were not detected for the complexes with tripod ligands unless they were coordinated to planar ligands such as salen or acacen. Studies of the interactions of cobalt complexes with amino acids demonstrated that tripod ligands promote the cysteine reaction, while the salen ligands promote histidine coordination, demonstrating a different mechanism of action. The results reported here are significant for better understanding and further design of zinc finger targeting compounds.
Assuntos
Cobalto/química , Complexos de Coordenação/química , Complexos de Coordenação/metabolismo , Dedos de Zinco , Aminas/química , Ligantes , Modelos MolecularesRESUMO
A systematic study of the influence of the first coordination sphere over the reactivity and structure of metallo-ß-lactamase (MßL) monozinc model complexes is reported. Three ZnII complexes with tripodal ligands forming the series [Zn(N-NNN)], [Zn(N-NNS)], and [Zn(N-NNO)] where N-NNX represents the tripodal donor atoms were investigated regarding their ability to mimic MßL. The tripodal series was inspired by MßL active sites in the respective subclasses, representing the (His, His, His) Zn1 site present in B1 and B3 subclasses, (His, His, Asp) present in the B3 subclass site and the thiolate present in B1 and B2 sites. The results were supported by electronic structure calculations. XAS analysis demonstrated that the ZnII electronic deficiency significantly changes in the order [Zn(N-NNS)] < [Zn(N-NNN)] < [Zn(N-NNO)]. This effect directly affects the reactivity over nitrocefin and amoxicillin, observed by the hydrolysis kinetics, which follows the same trend. NMR spectroscopy revealed the coordination of the carboxylic group in the substrate to the metal changes accordingly, affecting the hydrolysis kinetics. Our results also demonstrated that not only the Lewis acidity is changed by the ligand system but also the softness of the metal. [Zn(N-NNS)] is softened by the thiolate, promoting the ligand substitution reaction with solvents and favoring a secondary interaction with substrates, not observed for [Zn(N-NNO)]. XRD of the models reveals their similar geometric aspects in comparison to the crystal structure of GOB MßL. The present work demonstrates that the ZnII electronic details must be considered in the design of new MßL models that will further aid in the design of clinically useful inhibitors.