BPS2026 – Dynamic-structure redesign of calmodulin reduces Ca2+ leak and reveals mechanistic insights into RyR2 regulation

Calcium (Ca 2+ ) signaling underlies diverse physiological processes and is frequently dysregulated in disease. Calmodulin (CaM), the ubiquitous Ca 2+ sensor, regulates more than 300 protein targets, including the cardiac ryanodine receptor (RyR2)—a channel whose aberrant activity drives arrhythmias and heart failure. Owing to its extraordinary evolutionary conservation, CaM has long been considered “unimprovable,” with prior in silico engineering attempts yielding negligible or even reduced affinity compared to wild type. Here, we integrated insights across atomic, biochemical, and physiological scales to develop a holistic framework that incorporates both static- and dynamic-structure considerations, enabling computational redesign of CaM variants that not only enhance RyR2 binding but also improve its regulatory function. Redesigned CaMs displayed improved affinity for both the RyR2 peptide and the intact channel, with gains primarily due to slowed peptide dissociation mediated by the N-domain, as shown by steady-state and kinetic biochemical assays in vitro. Importantly, our strategy yielded improved RyR2 regulation ex vivo, evidenced by reduced Ca 2+ leak in isolated cardiomyocytes from both phosphomimetic S2808D and wild-type mouse models. Our results revealed that effective in silico CaM redesign requires not only stronger binding but also preservation of domain stabilization and the structural integrity of the RyR2 CaM-binding peptide. Structural alignment of publicly available CaM-RyR2 cryo-EM structures revealed peptide bending in PKA-phosphorylated models, while wild-type complexes maintained a straight conformation, further supporting our conclusion. Together, these findings demonstrate that effective CaM-mediated regulation of RyR2 emerges from a precise balance of affinity, domain stabilization, and dynamic integrity. Importantly, they show that even a highly conserved protein such as CaM can be computationally redesigned to restore physiological function at the cellular level in a disease-relevant model.