Investigation of non-corrin cobalt(II)-containing sites in protein structures of the Protein Data Bank

Abstract

Protein X-ray structures with non-corrin cobalt(II)-containing sites, either natural or substituting another native ion, were downloaded from the Protein Data Bank and explored to (i) describe which amino acids are involved in their first ligand shells and (ii) analyze cobalt(II)–donor bond lengths in comparison with previously reported target distances, CSD data and EXAFS data. The set of amino acids involved in CoIIbinding is similar to that observed for catalytic ZnIIsites,i.e.with a large fraction of carboxylate O atoms from aspartate and glutamate and aromatic N atoms from histidine. The computed CoII–donor bond lengths were found to depend strongly on structure resolution, an artifact previously detected for other metal–donor distances. Small corrections are suggested for the target bond lengths to the aromatic N atoms of histidines and the O atoms of water and hydroxide. The available target distance for cysteine (Scys) is confirmed; those for backbone O and other donors remain uncertain and should be handled with caution in refinement and modeling protocols. Finally, a relationship between both CoII—O bond lengths in bidentate carboxylates is quantified.