Synchronous and asynchronous transmitter release at nicotinic synapses are differentially regulated by postsynaptic PSD-95 proteins

Abstract

The rate and timing of information transfer at neuronal synapses are critical for determining synaptic efficacy and higher network function. Both synchronous and asynchronous neurotransmitter release shape the pattern of synaptic influences on a neuron. The PSD-95 family of postsynaptic scaffolding proteins, in addition to organizing postsynaptic components at glutamate synapses, acts transcellularly to regulate synchronous glutamate release. Here we show that PSD-95 family members at nicotinic synapses on chick ciliary ganglion neurons in culture execute multiple functions to enhance transmission. Together, endogenous PSD-95 and SAP102 in the postsynaptic cell appear to regulate transcellularly the synchronous release of transmitter from presynaptic terminals onto the neuron while stabilizing postsynaptic nicotinic receptor clusters under the release sites. Endogenous SAP97, in contrast, has no effect on receptor clusters but acts transcellularly from the postsynaptic cell through N-cadherin to enhance asynchronous release. These separate and parallel regulatory pathways allow postsynaptic scaffold proteins to dictate the pattern of cholinergic input a neuron receives; they also require balancing of PSD-95 protein levels to avoid disruptive competition that can occur through common binding domains.

Figure 1.

Simultaneous knockdown of PSD-95 and SAP102 reduces evoked EPSC amplitude in E8 CG neurons. Neurons transfected with the indicated constructs were patch clamped in the whole-cell configuration while stimulating presynaptic neurons with a bipolar extracellular stimulating electrode (arrows) to evoke maximal EPSCs. Dispersion of PDZ scaffolds with CRIPT reduced EPSC amplitude. RNAi knockdown of individual PSD-95 family members had no effect. Simultaneous knockdown of PSD-95 and SAP102, but not other combinations of PSD-95 family members, reduced EPSC amplitude. Values represent the mean ± SEM with n = 8–23 cultures per condition; n = 74 for controls. *p ≤ 0.05 by ANOVA with Dunnett's post hoc test.

Figure 2.

PSD-95/SAP102 knockdown alters α3*-nAChR expression and synaptic alignment. A, CG neurons transfected either with RFP as a negative control (Control, red) or RNAi targeting both PSD-95 and SAP102 (P95/S102-RNAi) and immunostained for surface α3*-nAChRs (green) and the presynaptic marker synaptotagmin (Syntag, blue). Scale bar, 10 μm. B, PSD-95/SAP102 knockdown reduced the level of α3*-nAChRs on the surface (left) and the mean size of α3*-nAChR clusters (right) on CG cell bodies, but did not alter the mean number of α3*-nAChR clusters aligned with presynaptic specializations (data not shown). C, Proximal neurites of CG neurons transfected with RFP as a negative control (red) or PSD-95/SAP102-RNAi and immunostained for surface α3*-nAChRs (green) and the presynaptic marker synaptotagmin (blue). Arrows indicate points of receptor/synaptotagmin alignment. Scale bar, 5 μm. D, PSD-95/SAP102 knockdown reduced the levels of α3*-nAChRs on the surface (left) and disrupted alignment of postsynaptic α3*-nAChR clusters under presynaptic specializations on proximal neurites (right), but did not alter the mean size of α3*-nAChR clusters (data not shown). E, Left, Whole-cell responses to 20 μm nicotine in a control neuron and a neuron transfected with PSD-95/SAP102-RNAi. Right, PSD-95/SAP102 knockdown decreased mean whole-cell responses to nicotine, whereas knockdown of SAP97 (S97-RNAi) had no effect. *p ≤ 0.05 by unpaired Student's t test (10–11 cultures/condition).

Figure 3.

PSD-95/SAP102 knockdown diminishes spontaneous mEPSC frequency and PPD. A, Left, Whole-cell patch-clamp recordings of mEPSCs in CG neurons in 1 μm TTX. Right, Overlay of 10 consecutive mEPSCs (gray) and the resulting mean (black) from the cells yielding the recordings. B, Knockdown of PSD-95 and SAP102 (P95/S102-RNAi), but not SAP97 (S97-RNAi), decreased mean mEPSC frequency in CG neurons (*p ≤ 0.05 by ANOVA with Dunnett's post hoc test). C, PSD-95/SAP102 knockdown also decreased mEPSC amplitude as seen in cumulative histograms (Cum Fraction; p ≤ 0.01 by K-S test; 9–15 cultures/condition). D, Paired responses in patch-clamped CG neurons exhibited PPD when presynaptic inputs were stimulated (arrows) at an interval of 20 ms. E, Neurons expressing PSD-95/SAP102-RNAi exhibited less PPD than did control neurons, whereas PPD was unchanged in neurons expressing SAP97-RNAi (*p ≤ 0.05 by unpaired Student's t test; 11–15 cultures/condition).

Figure 4.

SAP97 sustains nicotinic transmission during extended depolarizations and supports asynchronous ACh release. A, Top left, Traces showing mEPSCs recorded from CG neurons in the presence of 10 mm KCl and 1 μm TTX. Top right, Overlay of 10 consecutive mEPSCs (gray) and the resulting mean (black) from the cells yielding the recordings. Bottom, Under these conditions, SAP97 knockdown significantly decreased mEPSC frequency (left; *p ≤ 0.05 by ANOVA with Dunnett's post hoc test; 14–28 cultures/condition), but did not alter mEPSC amplitude (right). B, Top, Traces showing responses in CG neurons to asynchronous release of transmitter. The cells were transfected with the indicated RNAi constructs targeting individual PSD-95 family members. EPSCs that occurred ≥100 ms (dotted line) after the stimulus (arrows) were counted as asynchronous release events. Bottom, Knockdown of SAP97 decreased the mean number of asynchronous events that occurred in response to a single stimulus. Knockdown of PSD-95/SAP102 or PSD-95 alone did not alter asynchronous release (values presented relative to controls in same cultures; *p ≤ 0.05 by ANOVA with Dunnett's post hoc test; 4–19 cultures/condition).

Figure 5.

SAP97 knockdown impairs N-cadherin expression, a transmembrane protein important for asynchronous release. A, Left, CG neurons expressing RNAi constructs targeting PSD-95 family members (red) were immunostained for N-cadherin (green). Right, N-Cadherin (N-Cad) levels were reduced by SAP97 (S97-RNAi) but not by PSD-95 (P95-RNAi) or PSD-95/SAP102 (P95/S102-RNAi) knockdown (*p ≤ 0.05 by ANOVA with Dunnett's post hoc test; 5–17 cultures/condition). Scale bar, 5 μm. B, Disruption of N-cadherin interactions with a function blocking mAb (N-Cad mAb) selectively impaired asynchronous release. Left, Traces showing asynchronous release (as in Fig. 4B) with cells treated with IgG (control) or N-Cad mAb. Right, Quantification showing that the N-Cad mAb effect was equivalent to but not additive with the SAP-97-RNAi effect on asynchronous release (*p ≤ 0.05; **p ≤ 0.01 by ANOVA with Dunnett's post hoc test; 6 cultures/condition).

Figure 6.

Overexpression of SAP97 has a dominant-negative effect that depends on the GK domain. A, Left, Evoked EPSCs. Right, SAP97-GFP (S97-GFP) expression decreased evoked EPSC amplitude while SAP97/GK-GFP (S97-GK) did not (11–14 cultures/condition). B, Left, Whole-cell responses to rapidly applied nicotine. Right, SAP97-GFP expression diminished the mean whole-cell response to nicotine while SAP97/GK-GFP did not (4–10 cultures/condition). C, Left, Responses to paired stimuli. Right, SAP97-GFP expression decreased the magnitude of PPD (increased PPR), whereas SAP97/GK-GFP did not (8 cultures/condition). SAP97 acted as a dominant-negative via its GK domain to impair ACh release. *p ≤ 0.05 by ANOVA in A and C; ***p ≤ 0.001 by unpaired Student's t test in B.

Figure 7.

SAP102 requires a GK domain to enhance nicotinic synaptic input. A, Left, Evoked EPSCs. Right, SAP102-GFP (S102-GFP) increased the mean evoked EPSC amplitude while SAP102/GK-GFP (S102-GK) reduced it (6–9 cultures/condition). B, Left, Proximal neurites of transfected neurons (green) immunostained for surface α3*-nAChRs (red) and the presynaptic marker synaptotagmin (blue). Arrows indicate examples of receptor/synaptotagmin alignment. Right, SAP102-GFP expression increased the number of α3*-nAChR clusters aligned with presynaptic specializations (Syn α3*-nAChR Clust); SAP102/GK-GFP decreased it (4 cultures/condition). C, Left, Responses to paired stimuli. Right, SAP102-GFP increased PPD (decreased PPR) while SAP102/GK-GFP decreased PPD (increased PPR; 7 cultures/condition). D, Left, Whole-cell responses to nicotine. Right, SAP102-GFP did not alter whole-cell responses to nicotine (5 cultures/condition), but SAP102/GK-GFP decreased it (6 cultures/condition). *p ≤ 0.05; ***p ≤ 0.001 by unpaired Student's t test.