AMPA receptor nanoscale dynamic organization and synaptic plasticities

Highlights

• New imaging approaches have revealed the nanoscale organization of AMPA receptors.

• AMPA receptors are partly organized in nanodomains facing glutamate release sites.

• About half of AMPAR are not clustered and are mobile in synapses.

• AMPA receptor surface diffusion regulates post-synaptic short-term plasticity.

• AMPA receptor diffusion trapping at synapses participates to Long-Term Potentiation.

The emergence of new imaging techniques and molecular tools has refreshed our understanding of the principles of synaptic transmission and plasticity. Superresolution imaging and biosensors for measuring enzymatic activities in live neurons or neurotransmitter levels in the synaptic cleft are giving us an unprecedented integrated and nanoscale view on synaptic function. Excitatory synapses are now conceptualized as organized in subdomains, enriched with specific scaffolding proteins and glutamate receptors, molecularly organized with respect to the pre-synaptic source of glutamate. This new vision of basic synaptic transmission changes our understanding of the molecular modifications which sustain synaptic plasticities. Long-term potentiation can no longer be explained simply by an increase in receptor content at the synapse. We review here the latest data on the role of nanoscale and dynamic organization of AMPA type glutamate receptors on synaptic transmission at both basal state and during short and long-term plasticities.

Introduction

Since the emergence of the postulate that activity-dependent modulation of excitatory synaptic transmission could be a substrate of learning and memory [1], intense research has focused on understanding the molecular mechanisms through which neuronal activity could rapidly and endurably modify synaptic properties. Importantly, for any set of mechanisms to account for activity-driven synaptic plasticity, it has to have a fast (seconds) onset and sustain long-term changes. This is derived from experimental observations that robust changes in synaptic transmission can be induced by short tetanic activity and are maintained over hours and even days. Fast excitatory neurotransmission is mediated by glutamate, mostly occurs at small protrusions along dendrites – synaptic spines – and is almost exclusively mediated by AMPA subtype of ionotropic glutamate receptors (AMPAR). A major step forward in our understanding of the mechanisms of synaptic plasticity was achieved at the turn of the XXIst century, when studies from a number of laboratories proposed that rapid activity-driven changes in the number of AMPAR facing glutamate release sites could be the main substrate for modulations of synaptic efficacy [2,3]. The correlate hypothesis that AMPAR could enter and leave synapses within minutes to seconds came at a noticeable variance from the former view that synaptic receptor numbers were quite stable over time and that only changes in transmitter release or post-translational modifications of existing receptors could rapidly affect the efficacy of synaptic transmission [4].

Our current understanding for AMPAR dynamics is that they are trafficked through multiple pathways, including diffusion in the ER, directed intracellular vesicular transport on microtubules, exocytosis to the cell surface, endocytosis from the cell surface and random Brownian diffusion at the neuronal surface (in Ref. [4]) (Figure 1a). Each of these pathways can be modulated by neuronal activity, and thus likely contributes to regulate AMPAR numbers at synapses during activity dependent synaptic plasticities.

Setting AMPAR number at synapses results from a dynamic interplay between these trafficking pathways, the regulation of the total amount of AMPAR through biosynthesis and their postsynaptic stabilization. Surprisingly, despite major progress, the exact mechanisms through which AMPAR are reversibly stabilized at synapses and how this is controlled during various synaptic plasticities still remains elusive. This originates in part from the plethora of identified molecular AMPAR interactors. First and foremost, AMPAR form a complex with a variety of auxiliary subunits [5^•] who all seem to regulate AMPAR function. In turn, each of the AMPAR pore-forming and auxiliary subunits can bind a variety of intracellular scaffold elements, and possibly extracellular matrix and adhesion proteins (e.g. and in Refs. [6,7]). Most of these interactions are regulated by neuronal activity and are thus candidates for reversible synaptic stabilization of AMPAR. While modulation of both AMPAR stabilization and trafficking processes seem to be involved in synaptic plasticity events, the respective exact contribution of each modality is still relatively mysterious.

In addition to the mere number of synaptic AMPAR, their precise localization with respect to glutamate release sites is a determinant factor of the efficacy of synaptic transmission. Indeed, AMPAR have a low affinity for glutamate and are thus only efficiently activated when located in the post-synaptic membrane within less than ∼100 nm of pre-synaptic glutamate release sites. In addition, sites of AMPAR trafficking between the cell surface and intracellular compartments are distributed all along the neuronal membrane and are not particularly localized at synapses. Hence, AMPAR surface diffusion is likely to play a key role both as a shuttling system between synaptic sites and extrasynaptic sites of recycling, and as a mechanism for rapid entry and exit of receptors from an area where they can be activated by transmitter release. We will review here the recent studies that have facilitated our understanding of the relationship between AMPAR dynamics, their nanoscale organization and synaptic plasticities, and highlight the outstanding questions.