Structure/calcium affinity relationships of site III of calmodulin: testing the acid pair hypothesis using calmodulin...

Calmodulin mutants in which the calcium binding affinity of site IV was greatly reduced by a D133E mutation were prepared using site-specific, cassette-mediated mutagenesis as a multisite calcium binding protein model to examine structure/calcium affinity relationships in site III of calmodulin. Tryptophan was introduced in position 92 of the calmodulin mutants as a fluorescent label to monitor the calcium-induced structural changes in the C-terminal domain of calmodulin. The five calmodulin mutants, 3xCaM, 3zCaM, 4xCaM, 4zCaM, and 4xzCaM, were designed so that there were three or four acidic amino acid residues in chelating positions of site III with acid pairs on either the X and/or Z coordinating axes. The calcium dissociation constant of site III, KIII, of the five calmodulin mutants changes in a descending order from 3xCaM (237 microM), 3zCaM (140 microM), 4xCaM (5.8 microM), 4zCaM (3 microM), to 4xzCaM (2 microM), and these KIII values are significantly lower than that of F92W/D133E calmodulin (335 microM) in which three acidic residues with no acid pairs were present in site III [Wu, X., & Reid, R. E. (1997) Biochemistry 36, 3608-3616]. These results indicate that the calcium affinity of site III increases when the number of the acidic chelating residues increases from three to four, when the number of acid pairs increases from zero to one and further to two, and when the location of the acid pair is changed from the X axis to the Z axis. This study provides the first evidence that the acid pair hypothesis which correlates the nature of the chelating residues with the calcium affinity of the hlh motif is applicable to a multisite calcium binding protein model. The Hill coefficients indicate that reversal of the sequence of filling of the calcium binding sites in the C-terminal domain from IV --> III to III --> IV also changes the site cooperativity from positive to negative. The cooperativity returns to positive when the proteins are titrated in the presence of a calmodulin-binding peptide. Data from the present study also demonstrate that calmodulin mutants with a decreased calcium affinity have a reduced efficiency in phosphodiesterase regulation at low calcium concentrations (50 microM). However, high calcium concentrations (15 mM) restore the phosphodiesterase regulatory activity of the calmodulin mutants to a level obtained with F92W calmodulin, indicating that the mutations alter calcium regulation of calmodulin-mediated phosphodiesterase activity without affecting the interaction between calmodulin and the enzyme.