Metal selectivity in the E. coli Nickel(II)- and Cobalt(II)-responsive transcriptional regulator, RcnR

Cellular trafficking systems for transition metals require importers, exporters, chaperones and regulators that respond to specific metals. Understanding how different metals are recognized by proteins in order to generate specific biological responses is a goal of this research. E. coli RcnR is a 40.4 kDa tetrameric transcriptional repressor that responds to the binding of Ni(II) or Co(II) at micromolar concentrations to allow the expression of the exporter, RcnA. E. coli RcnR has significant amino acid sequence homology with M. tuberculosis CsoR, a Cu(I) responsive regulatory protein. RcnR and CsoR are members of a new structural class of transcriptional regulators characterized by an &agr;-helical structure consisting of a four-helix bundle. RcnR binds to a variety of metals in vitro, and it is thought that the metal ion-selective biological response derives from both preferences in ligand selection and coordination number. This model is similar to that seen for E. coli NikR, the Ni-responsive transcription regulator responsible for the repression of the nickel importer, NikABCDE. Prior data obtained from XAS show that RcnR forms six-coordinate complexes with its cognate metal ions (Ni(II) and Co(II)) that have M(N/O)5S ligand donor atom sets. Mutagenesis and lacZ data suggests that these ligands are derived from the coordination of the N-terminal amine and the side chains of Cys35, His3, His60 and His64. The largest structural difference between the Co(II) and Ni(II) complexes was found to be the M-Scys35 distance (2.31 Å and 2.62 Å for Co(II) and Ni(II), respectively). Metal substitutes of both wild-type and mutant RcnR proteins (A2*, H3L, H3C, H3E, H60C, H64C and H67C) have been prepared and characterized. The metal complexes were assayed for RcnA de-repression using LacZ reporter assays. The data from the wild-type metal complexes with RcnR show that metal responsiveness in RcnR is linked to the coordination number and geometry of the metal ions; the cognate metal ions Ni(II) and Co(II) are six-coordinate while the non-cognate metals have low coordination numbers. All the metals bind at the same locus, involving C35, the only cysteine residue in the protein. However, only binding the six-coordinate metal ions, Ni(II) and Co(II), results in de-repression of RcnA. The role of specific ligands in creating the metal binding site was addressed by mutagenesis and XAS. These studies show that the binding of the N-terminal amine is important in discriminating cognate from non-cognate metals. Further, Ni(II) and Co(II) are recognized differently; the complexes formed have distinct interactions with His3 and very different M-SCys35 distances. Mutation of His3 to aspartate resulted in a protein that now responds to the binding of Zn(II) ions, to our knowledge the first example of changing the metal-specificity of a transcriptional regulator