Two-dimensional NMR Studies of Selenomethionyl Calmodulin

Abstract

Calmodulin (CaM) is a ubiquitous calcium regulatory protein that can interact with almost 30 different target proteins. The majority of the CaM-binding domains of the target proteins are believed to interact with two hydrophobic surfaces on Ca2+-CaM; these two regions are very rich in Met residues. To obtain more information about the role of these residues, we have biosynthetically incorporated selenomethionine (SeMet) in place of the nine Met residues of CaM. Amino acid analysis shows that the SeMet-CaM contains 15% Met and 85% SeMet. SeMet-CaM retains many of the properties of the wild-type protein; it activates the enzyme cyclic nucleotide phosphodiesterase, it binds to phenyl-Sepharose and myosin light chain kinase (MLCK) in a calcium-dependent manner, and it experience a calcium-dependent band shift during SDS-gel electrophoresis. Moreover, by comparing the natural abundance (1H, 13C)-heteronuclear multiple quantum coherence (HMQC) spectra of the calcium, apo and target peptide-bound forms of wild-type CaM and SeMet-CaM, we have found that the two proteins have very similar, if not identical, structures. Thus, the substitution of SeMet for Met does not cause a change in the conformation and function of CaM, in agreement with the results obtained for other proteins. The apo, calcium and target peptide-bound forms of SeMet-CaM were subsequently studied by natural abundance (1H, 77 Se)-heteronuclear multiple bond correlation (HMBC) and (1H, 13C)-HMQC NMR. Nine well-resolved 77Se resonances could be observed. Substitution of SeMet for Met gave rise to the same 1H and 13C chemical shift changes for each individual Met residue, this facilitated making the assignments from known 1 H, 13C assignments of the Met residues. Some of these assignments were confirmed by studying Met→Leu mutants of CaM. With the exception of Met76, which always remains solvent exposed, all resonances experienced large 77Se chemical shift changes upon the addition of Ca2+ and the MLCK peptide. The large shift changes indicate that the electron distribution in the SeMet side-chain can be adjusted for the different states of CaM, suggesting that the polarizability of sulfur or selenium may be important for the proper functioning of CaM. This study also shows that the natural abundance (1H, 77Se)-HMBC experiment provides a sensitive approach for the study of SeMet protein