PROJECT SUMMARY Learning, cognition, and memory require dynamic remodeling of hippocampal synapses, which in turn requires Ca2+/calmodulin-dependent kinase II (CaMKII). CaMKII mediates two opposing modes of synaptic plasticity, long term potentiation (LTP) and depression (LTD), that are induced by distinct Ca2+ stimuli. Both low and high Ca2+ induce CaMKII autophosphorylation (p) at T286, that is required for both LTP and LTD. Additionally, LTP requires CaMKII binding to the NMDA receptor subunit, GluN2B, during high [Ca2+] while LTD requires CaMKII autophosphorylation at T305/306 during low [Ca2+]. Further, these three mechanisms can undergo complex cross-regulation which requires the CaMKII 12-meric holoenzyme. Interestingly, pT286 positively regulates both GluN2B binding and pT305/306 while GluN2B binding and pT305/306 are mutually exclusive. It is unknown how these reactions and interactions are spatiotemporally encoded within holoenzymes and thus how LTP versus LTD signal computation is accomplished by CaMKII. For example, it has been shown that pT286 must occur between two neighboring kinase domains in the holoenzyme. It is still unclear what determines a functional kinase domain neighbor. Moreover, in vitro binding studies have shown that CaMKII holoenzymes are required for binding to GluN2B, suggesting that this interaction may require multiple subunits. Still, it is unknown what is the required stoichiometry and subunit geometry required for CaMKII-GluN2B binding. Initial results suggest that the holoenzyme rules for pT286 and GluN2B binding are fundamentally different. Therefore, this proposal will investigate my hypothesis that LTP versus LTD mechanisms are regulated by structurally distinct features within CaMKII holoenzymes. The approach will utilize several CaMKII “structural mutants” that have disrupted holoenzyme structure (hexamers, dimers, and monomers). These mutants will be used as tools to define the spatiotemporal dynamics of autophosphorylation within holoenzymes and the subunit stoichiometry and geometry required for GluN2B binding. The results of this proposal will provide insight into how molecular signal computation underlying the LTP versus LTD decision is encoded within CaMKII holoenzymes.