Edinburgh Research Explorer

Presynaptic NMDA receptors differentially regulate release probability at neocortical excitatory synapses

Research output: Contribution to conferencePoster

Related Edinburgh Organisations

Original languageEnglish
Publication statusPublished - 14 Oct 2012
EventSociety for Neuroscience Annual Meeting 2012 - New Orleans
Duration: 13 Oct 201217 Oct 2012

Conference

ConferenceSociety for Neuroscience Annual Meeting 2012
CityNew Orleans
Period13/10/1217/10/12

Abstract

Classically, NMDA receptors are postsynaptic coincidence detectors in plasticity, but recent studies suggest that they also exist presynaptically (preNMDARs), where they have been implicated in regulating short-term plasticity. However, precisely how preNMDARs control release is poorly described.
We previously found that, in mouse visual cortex layer 5, preNMDARs control neurotransmitter release at connections from pyramidal cells (PCs) to other PCs, to somatostatin (SOM) interneurons (INs), and to a subset of non-basket-cell parvalbumin (PV) INs, while specifically leaving those to basket cells unaffected. Here, we combine computer modeling and 2-photon imaging to investigate the impact further. We tuned the Tsodyks-Markram short-term plasticity model to 30-Hz EPSP trains at these three excitatory synapses before and after preNMDAR blockade (200 µM D/L-AP5), and found that it consistently reduced baseline release probability (PC: 67±3%, n=7; SOM: 45±2%, n=9; PV: 71±5%, n=9; all: p<0.001). However, AP5 did not affect the time constants of depression (PC: p=0.63; SOM: p=0.38) or facilitation ( PC: p=0.15; SOM: p=0.65; PV: p=0.2), nor the facilitation constant (PC: p=0.15; SOM: p=0.67; PV: p=0.38). Onto PV, however, AP5 did increase the depression time constant (250% ± 10%, n=9; p<0.001). In keeping with reduced release probability, AP5 consistently increased failure rates at PC inputs to the three cell types (PCs: 268%±809% of baseline, n=9; p<0.001; SOM: 238%±715%, n=8; p<0.01; PV: 349%±242%, n=3; p<0.05).
The preNMDAR-mediated increase in baseline release probability may be linked to bouton calcium signals. To test this idea, we 2-photon imaged calcium in PC boutons. Although spike-evoked calcium transients were surprisingly unaffected by AP5 wash-in (100±11%, n=5, p=0.84 vs. controls), 5 of 12 boutons showed significant changes in basal calcium (63±4%, p<0.001), consistent with the model. We thus extended the model to include preNMDAR control of STP. This extension predicted a sigmoidal rate dependence of AP5 blockade -- with an 8.4Hz midpoint -- that we verified experimentally.
Our results suggest that preNMDARs specifically regulate release probability at a subset of synapses from PCs by locally controlling bouton resting calcium levels. PreNMDARs are thus well positioned to control the flow of information in the cortical microcircuit.

Event

Society for Neuroscience Annual Meeting 2012

13/10/1217/10/12

New Orleans

Event: Conference

ID: 14722470