Identification of Noninterfacial Amino Acids Important for Molecular Recognition in Calmodulin

Abstract

Calmodulin (CaM) contains two structurally homologous domains that cooperatively bind to a range of different target proteins, such that upon binding the opposing N- and C-domains wrap around the CaM-binding sequence. The N-domain has a binding affinity (Kd = 24 µM) that is approximately 3-orders of magnitude weaker than the C-domain (Kd = 11 nM) to the plasma membrane Ca-ATPase (PMCA). These large differences in binding affinities facilitate ordered binding, which is necessary for the productive activation of many target proteins. To better understand design principles that facilitate molecular recognition, we have used directed evolution combined with yeast surface display to identify mutations that enhance binding between the N-domain of CaM and the PMCA, permitting the identification of combinatorial families of mutate N-domain proteins with nanomolar binding affinities (similar to that of the C-domain). All observed mutations occur at noninterfacial sites that are naturally variable in Nature, suggesting that hypervariable sites between different species may fine tune binding affinities. Mutations in the N-domain that selectively destabilize the unbound state commonly result in enhanced binding affinities with other CaM-binding sequences (i.e., skeletal myosin light chain kinase and ryanodine receptor). In contrast, mutations that structurally couple with residues in the binding interface result in selective high-affinity binding to the PMCA; these mutations commonly result in decreased binding to other target proteins. In total, these results indicate the value of using directed evolution approaches to identify underlying principles that determine binding affinities.This research was supported by the Defense Threat Reduction Agency (DTRA).