The Delicate Bistability of CaMKII Activation

Abstract

Calcium/calmodulin-dependent protein kinase II (CaMKII) is one of the most abundant proteins in the brain and is essential for learning and memory. The activity of CaMKII is regulated both by calmodulin binding and by phosphorylation on Thr286, which maintains the kinase in a partially active state even in the absence of a calcium/calmodulin signal, thus prolonging the effects of transient calcium signaling. CaMKII exists as a 12 subunit holoenzyme, and phosphorylation at Thr286 occurs by an intra-holoenzyme, inter-subunit reaction. An autophosphorylating kinase, in conjunction with a phosphatase, can potentially form a bistable switch, and it has been proposed that bistability in the CaMKII system may constitute the biochemical change underlying long-term memory. Previous modeling efforts have suggested that bistability is likely under physiological conditions, but experimental studies have proved inconclusive. Previous modeling efforts involved several significant approximations in order to overcome the combinatorial complexity inherent in a multi-subunit, multi-state system. Here we develop a stochastic, particle-based model of CaMKII dynamics and activation which naturally avoids the issues of combinatorial complexity, and thus allows us to study an exact model without resorting to severe approximations. We find that bistability is possible, but for any reasonable choice of parameters bistability only occurs at calcium concentrations much higher than basal calcium levels, and then only over a very narrow range of calcium concentration. We conclude that bistability is not a physiologically relevant feature of the CaMKII system, which should appear to behave as an ultra-sensitive switch. On the other hand, the system dynamics are generally very slow. Transiently activated kinase can maintain its activity over the time scales of many of the published experimental protocols, which may account for the conflicting reports of kinase bistability. This work was supported by NIH grants 1F32-NS077751-01 and P41GM103313.