Bidirectional synaptic plasticity as a consequence of interdependent Ca2+-controlled phosphorylation and dephosphorylation pathways.

Abstract

Postsynaptic Ca2+ signals of different amplitudes and durations are able to induce either long-lasting potentiation (LPT) or depression (LTD). The bidirectional character of synaptic plasticity may result at least in part from an increased or decreased responsiveness of the glutamatergic alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor (AMPA-R) due to the modification of conductance and/or channel number, and controlled by the balance between the activities of phosphorylation and dephosphorylation pathways. AMPA-R depression can be induced by a long-lived Ca2+ signal of moderate amplitude favouring the activation of the dephosphorylation pathway, whereas a shorter but higher Ca2+ signal would induce AMPA-R potentiation resulting from the preferential activation of the phosphorylation pathway. Within the framework of a model involving calcium/calmodulin-dependent protein kinase II (CaMKII), calcineurin (PP2B) and type 1 protein phosphatase (PP1), we aimed at delineating the conditions allowing a biphasic U-shaped relationship between AMPA-R and Ca2+ signal amplitude, and thus bidirectional plasticity. Our theoretical analysis shows that such a property may be observed if the phosphorylation pathway: (i) displays higher cooperativity in its Ca2+-dependence than the dephosphorylation pathway; (ii) displays a basal Ca2+-independent activity; or (iii) is directly inhibited by the dephosphorylation pathway. Because the experimentally observed inactivation of CaMKII by PP1 accounts for this latter characteristic, we aimed at verifying whether a realistic model using reported parameters values can simulate the induction of either LTP or LTD, depending on the time and amplitude characteristics of the Ca2+ signal. Our simulations demonstrate that the experimentally observed bidirectional nature of Ca2+-dependent synaptic plasticity could be the consequence of the PP1-mediated inactivation of CaMKII.