Figure 1. Model Structure.

(A) Overview of model, indicating key trafficking steps for AMPAR and CaMKII.

(B–H) Chemical reaction schemes for pathways in model. Curved lines with arrows are enzymatic reactions catalyzed by molecules at the curves. Straight lines represent binding or unimolecular reactions.

(B) Details of AMPAR model. The modelled AMPAR is a tetramer with two subunits each of GluR1 (circles) and GluR2 (triangles). There are 16 phosphorylation states each in the cytosol and synaptic membrane. These are represented in expanded form in the lower portion of (B), which shows the internalised pools of receptors. Black filling of the left half of the GluR1 circle indicates phosphorylation of Ser845, and of the right half indicates phosphorylation of Ser831. Endocytosis occurs for the receptors with no GluR1-Ser845 phosphorylation, and exocytosis and degradation occur for the receptors with both GluR1 subunits phosphorylated on the Ser845 site. Exchange of receptors with the bulk AMPAR pool occurs for the unphosphorylated state only, outlined in black.

(C) CaMKII model. The dashed line for phosphorylation of CaMKII–PSD is applicable only for the bistable CaMKII models described in Figure 7.

(D) CaM activation.

(E) PP1 activation.

(F) PP2B (calcineurin) activation.

(G) cAMP formation. The unstimulated phosphodiesterase molecules (PDEs) also degrade cAMP, but at a lower rate than the activated forms illustrated. In the cAMP model we include diffusive exchange of cAMP with a dendritic compartment.

(H) PKA activation.

AMP, adenosine monophosphate; ATP, adenosine triphosphate; I1, inhibitor of PP1; Ng, neurogranin; PKA_inhib, inhibitor of PKA; PKC, protein kinase C; PP2A, protein phosphatase 2A.