Structural and biochemical insights into mammalian cobalt-substituted methionine sulfoxide reductase B1 using UV-visible spectroscopy and high-resolution NMR spectroscopy
MetadataShow full item record
Prior to this study it was discovered that MsrB1 from Mus musculus expressed in Escherichia Coli binds cobalt(II) (hereafter cobalt) in cobalt-supplemented growth media, and it had further been demonstrated that the His-tag was not responsible for this metal uptake. The aim of this study was to investigate the effects of cobalt on the growth of E. Coli in culture, characterize the metal-uptake and metal-binding site of cobalt-substituted MsrB1 by UV-visible spectroscopy and to gain structural information about the protein by high resolution NMR spectroscopy.The effects of cobalt on growth of E. Coli were studied by growing cultures in Lysogeny broth (LB) and minimal (M9)-media supplemented with different concentrations of cobaltdichloride (CoCl2) and monitoring culture growth by optical density (OD) measurements. Growth rates were found to decrease with increasing concentrations of CoCl2. To study the cobalt-uptake of MsrB1, the protein was recombinantly expressed in E. Coli in different cobalt-supplemented growth media, and the purified protein analyzed by UV-vis spectroscopy. Cobalt-uptake was demonstrated in all cases by characteristic absorption peaks owing to the cobalt-ligand complex, and the wavelengths of these peaks matched those of a tetrahedral four-coordinated cobalt-Cys complex. It was argued that these four Cys residues should be the same that constitute the zinc(II)-binding site of MsrB1, indicating that cobalt simply replaces zinc as a structural metal ion in cobalt-substituted MsrB1 (Co-MsrB1). It was argued that the intracellular concentration of cobalt in E. Coli should be significantly higher than zinc, and that this together with similar ionic radii for cobalt and zinc leads to the formation of Co-MsrB1. MsrB1 expressed in nickel(II)-supplemented LB did not lead to formation of Ni-MsrB1, which was argued to result from nickel not being released directly into the cytosol in E. Coli. Co-MsrB1 was produced by zinc-starvation of E. Coli followed by expressing the protein in zinc-free minimal medium supplemented with CoCl2. To investigate if pH-titratable groups could be detected, the protein was dialyzed against buffers with pH 4.9-11.5 and the molar extinction coefficient was found from UV-vis absorption spectra in the different pH. Two titration curves were observed, but assignment of the titrations to specific residues could not be made. Further, Co- and Zn-MsrB1 was dialyzed against buffers with metal chelating agents to remove the metal ions from the two proteins. Cobalt was successfully removed at pH 5.0 and 5.5, while removal of zinc from Zn-MsrB1 was not detected, demonstrating that zinc is more tightly bound to the Cys-ligands than cobalt.To study the ratio of formation of Co-MsrB1 and the native zinc-form Zn-MsrB1 in CoCl2-supplemented growth media, the molar extinction coefficient of Co-MsrB1 was determined, and the concentration of Co-MsrB1 in purified protein samples from protein expression in different growth media was determined. The Co-MsrB1:Zn-MsrB1 ratio was found to be 0.2 in M9 medium supplemented with 10 µM CoCl2, 0.1 in LB supplemented with 50 µM CoCl2 and 0.03 in LB supplemented with 10 µM CoCl2. From 2D- and 3D-NMR experiments on 13C- and 15N-enriched Co-MsrB1, a 70 % backbone assignment and 50 % side chain-assignment was accomplished using Computer Aided Resonance Assignment (CARA). The four structural Cys-residues of the native protein was not found, while the other three Cys-residues of MsrB1 were assigned, confirming that the same Cys-residues are responsible for coordination of cobalt and zinc in MsrB1. Many strongly shifted signals were observed in 1D 1H spectra of Co-MsrB1, some as far upfield as 350 ppm and downfield as -80 ppm, and it was argued that most of the unassigned residues should be found outside the spectral width of the 2D and 3D-NMR spectra.To gain further structural information about Co-MsrB1, pseudocontact shifts (PCSs) were determined for the assigned HN-atoms of Co-MsrB1 by using the published chemical shifts of for Zn-MsrB1. The PCSs were analyzed by AnisoFit using three conformers and the mean conformer of the published Zn-MsrB1 structure, and the best correlations between observed and calculated PCSs were found for conformer 3. The PCSs were plotted against their hypothetical distance to cobalt using the structure of Zn-MsrB1, and a very good PCS-distance proportionality was found, indicating that Co-MsrB1 and Zn-MsrB1 have the same overall fold and structure. The AnisoFit calculations and PCSs of the N-terminus suggested that the N-terminus spends significant time in the proximity of the metal-binding site, and it was argued that this proximity ensures high catalytic efficiency due to the short distance between the catalytic and resolving Cys-residues.