4.1N and SAP97 regulate different phases of AMPA receptor intracellular transport

Abstract

Changes in the number of synaptic AMPA subtypes of glutamate receptors (AMPAR) underlie many forms of synaptic plasticity. These variations are controlled by a complex interplay between their intracellular transport (IT), export to the plasma membrane, stabilization at synaptic sites, and recycling. The differential molecular mechanisms involved in these various trafficking pathways and their regulation remains partly unknown. We have recently reported the visualization of AMPAR IT in cultured hippocampal neurons and demonstrated its regulation during synaptic plasticity inducing protocols (Hangen, Cordelieres et al., 2018), opening the path to the differential analysis of the mechanisms controlling AMPAR transport and exocytosis. The cytosolic C-terminal (C-ter.) domain of AMPAR GluA1 subunit is specifically associated with cytoplasmic proteins that could be implicated in the regulation of their IT such as 4.1N and SAP97. Here we analyze how interactions between GluA1 and 4.1N or SAP97 regulate IT and exocytosis at the plasma membrane in basal condition and after cLTP induction. We use sh-RNA against 4.1N and SAP97 and specific mutations and deletions of GluA1 C-ter. domain to characterize how these interactions are involved in coupling AMPAR to the transport machinery. The down-regulation of both 4.1N or SAP97 by shRNAs decrease GluA1 containing vesicle number, modify their transport properties and decrease GluA1 export to the PM, indicating their role in GluA1 IT. The total deletion of the C-ter. domain of GluA1 fully suppresses its IT. Disruption of GluA1 binding to 4.1N decreases the number of GluA1 containing transport vesicles, inhibits GluA1 externalization but does not affect the transport properties of the remaining GluA1 containing vesicles. This indicates a role of the 4.1N-GluA1 interaction during exocytosis of the receptor in basal transmission. In contrast, disrupting the binding between SAP97 and GluA1 modifies the basal transport properties of GluA1 containing vesicles and decreases GluA1 export to the plasma membrane. Importantly, disrupting GluA1 interaction with either 4.1N or SAP97 prevents both the cLTP induced increase in the number of GluA1 containing vesicles observed in control and GluA1 externalization. Our results demonstrate that specific interactions between 4.1N or SAP97 with GluA1 have different roles in GluA1 IT and exocytosis. During basal transmission, the binding of 4.1N to GluA1 allows the fusion/fission membrane exocytosis whereas the interaction with SAP97 is essential for GluA1 IT. During cLTP the interaction of 4.1N with GluA1 allows both IT and exocytosis of the receptor in hippocampal cultured neurons. Altogether, our results identify the differential roles of 4.1N and SAP97 in the control of various phases of GluA1 IT.