Differential stimulation of NAD kinase and binding of peptide substrates by wild-type and mutant plant calmodulin isoforms

Calmodulin from Arabidopsis thaliana consists of at least four isoforms, which differ in their deduced amino acid sequences by as many as six conservative substitutions. To determine whether these differences are biochemically significant, cDNAs encoding three of the four isoforms were engineered to produce recombinant proteins in Escherichia coli, purified to apparent homogeneity, and assayed for their abilities to activate pea leaf NAD kinase in vitro. The CaM-2 isoform was a significantly more efficient activator of NAD kinase compared with the CaM-4 and -6 isoforms based on both the apparent V(max) it elicited and the K_0.5 activation. These results are consistent with the hypothesis that the Arabidopsis CaM isoforms have evolved to optimize the protein's interaction with different Ca²⁺/CaM-regulated target enzymes. The ability to activate NAD kinase was also investigated for a carboxy-terminal nonsense mutant of CaM-6 (CaM-6M), which substituted 14 hydrophilic amino acids for a region of seven amino acids that normally form an exposed hydrophobic surface when wt CaM-6 binds Ca²⁺. CaM-6M-activated NAD kinase displayed an apparent V(max) that was reduced 40% and a K_0.5 that was an order of magnitude greater than the CaM-6-activated enzyme. The Ca²⁺-dependence of CaM-6M to activate NAD kinase was identical to that of CaM-6, but CaM-6M bound synthetic peptide substrates with lower apparent affinity than did CaM-6. Thus, the carboxy-terminal hydrophobic domain of CaM appears to be critical for its interaction with NAD kinase. In contrast, amino-terminal fusions of a hydrophilic, α-helical 12-residue c-myc epitope tag to CaM-2 and -4 yielded proteins that activated NAD kinase to apparent V(max) values within 10% of those obtained with the wild-type CaM proteins.