Long-term depression-associated signaling is required for an in vitro model of NMDA receptor-dependent synapse pruning

Abstract

Activity-dependent pruning of synaptic contacts plays a critical role in shaping neuronal circuitry in response to the environment during postnatal brain development. Although there is compelling evidence that shrinkage of dendritic spines coincides with synaptic long-term depression (LTD), and that LTD is accompanied by synapse loss, whether NMDA receptor (NMDAR)-dependent LTD is a required step in the progression toward synapse pruning is still unknown. Using repeated applications of NMDA to induce LTD in dissociated rat neuronal cultures, we found that synapse density, as measured by colocalization of fluorescent markers for pre- and postsynaptic structures, was decreased irrespective of the presynaptic marker used, post-treatment recovery time, and the dendritic location of synapses. Consistent with previous studies, we found that synapse loss could occur without apparent net spine loss or cell death. Furthermore, synapse loss was unlikely to require direct contact with microglia, as the number of these cells was minimal in our culture preparations. Supporting a model by which NMDAR-LTD is required for synapse loss, the effect of NMDA on fluorescence colocalization was prevented by phosphatase and caspase inhibitors. In addition, gene transcription and protein translation also appeared to be required for loss of putative synapses. These data support the idea that NMDAR-dependent LTD is a required step in synapse pruning and contribute to our understanding of the basic mechanisms of this developmental process.

Keywords: Caspase; Confocal microscopy; Dendritic spine; Development; Neuron culture; Phosphatase; Protein synthesis; Synapse; Transcription.

Figure 1. Experimental design

(A) Dissociated neuronal cultures at 18 – 22 days DIV have fully-branched dendrites studded with mature spines that can be visualized with fluorescent markers using confocal microscopy, and the changes in synapse density can then be quantified. Scale bar, 20 μm. (B) High magnification confocal images of dendrites with spiny synapses. Synapsin antibody staining (red) in the presynaptic terminal of Neuron A is apposed to the GFP-labeled spines and dendrite (green) of postsynaptic Neuron B. NMDA treatment of dissociated neuronal cultures causes a decrease in frequency of miniature EPSCs. (C) Representative traces from mEPSCs recorded from cultured neurons pre- and post-NMDA treatment. (D) Data from cells, normalized to baseline, after NMDA treatment shows a decrease in mEPSC frequency, but no change in amplitude (n = 10 cells, 3 biological samples). Paired analyses of mEPSC recordings from individual cells pre-NMDA and post-NMDA treatments: (E) frequency and (F) amplitude. Some points/error bars are obscured by overlying symbols with similar values. * p _freq = 0.031. (G) Schematic diagram of experimental timeline showing that dissociated neurons were cultured for 21 days, infected with a modified Sindbis-eGFP to fill some of the cells, treated twice with the LTD-inducing drug, NMDA (20μM for 3 minutes), allowed to recover two or four hours, and fixed. Cells were then stained for synapse markers (immunofluorescence), imaged, and putative synapse numbers determined.

Figure 2. NMDA-induced LTD results in synapse elimination

(A) Confocal images of neurons after treatments with media alone (CTRL), or one (1X NMDA) or two (2X NMDA) treatments with NMDA showing fluorescent immunoreactive staining of the presynaptic bouton (red, synapsin) and postsynaptic spine (green, GFP) in 21 DIV cultured dissociated neurons. Scale bar, 5 μm. (B1) Synapse loss occurs reliably after two NMDA treatments in dissociated neurons. Quantification of synapse density (averaged means of ^# of synapses/20 μm dendrite) in mock-treated (CTRL), once-treated (1X NMDA), and twice-treated (2X NMDA) cultured cells. (B2) Quantification of normalized synapse density (% of CTRL) in mock-treated (CTRL), once-treated (1X NMDA), and twice-treated (2X NMDA) cultured cells. Data show that synapse density changes at 2 hrs or 4 hrs post-NMDA treatment are similar. N = at least three independent biological samples. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): * p < 0.05, ** p < 0.01, *** p < 0.001; ns, not significant. (C1) NMDA-mediated synapse loss occurs similarly at both proximal and distal synapses. Representative confocal image of a GFP-expressing neuron, with 20 μm-boxed regions marking examples of proximal (20 – 50 μm from the soma) and distal (> 100 μm from the soma) dendritic areas used in analyses shown in (C2). Scale bar, 20 μm. (C2) Quantification of normalized synapse density for proximal and distal synapses in mock-treated (CTRL) and treated (NMDA) cells. Data are averaged means of synapses from at least three biological samples. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance: *** p < 0.001; ns, not significant.

Figure 3. NMDA-induced synapse loss is not likely to be due to loss of synapsin or GFP

NMDA-induced synapse loss is still observed using an alternative presynaptic marker, bassoon (A1), or postsynaptic marker, PSD-95 (A2). Confocal images after 0 (CTRL) and 2 NMDA treatments showing fluorescent immunoreactive staining using alternative markers of (A1) the presynaptic bouton (red, bassoon), and (A2) the postsynaptic spine, PSD-95. Scale bar, 5 μm. (A3) Quantification of normalized synapse density from neurons stained with alternative antibody combinations, anti-GFP/anti-bassoon, and anti-PSD95/anti-synapsin. Data are averaged means of synapses from at least three independent experiments. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): *** p < 0.001. (B1) Synapse pruning is unlikely to be a result of microglia action. Example image of an Iba1-positive microglial cell contacting a GFP-expressing neuron. Scale bar, 20 μm. (B2) Relative Iba-1-immunoreactivity in neuronal cultures using our neuron culture protocol (left), and microglia-enriched cultures grown in serum-containing media (Iba1-positive control, right). Scale bars, 50 μm.

Figure 4. NMDA treatment results in spine loss

(A1) Data are averaged means of spine density (^# spines/20 μm dendrite) in cultured neurons after mock or NMDA treatment from 16 independent experiments. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): * p = 0.025. (A2) Spine loss is not correlated with putative synapse loss. Data are averaged means of spine and synapse numbers, expressed as % spine loss and % synapse loss, in cultured neurons after mock or NMDA treatment from 16 independent experiments. R² = 0.014; p = 0.636. (B1) Confocal image (top; scale bar, 2 μm) and surface rendering (bottom; scale bar, 1 μm) of spiny dendritic sections (GFP-expressing) with punctate antibody staining to endogenous synapsin protein. Arrows indicate examples of ‘unpaired spines’. Circle marks a 2-μm radius region of interest for an ‘unpaired spine with synapsin close’. Center of the white circle is on an unpaired spine head center; any unpaired synapsin within the circle denotes ‘synapsin close’. (B2) NMDA treatment increases the number of unpaired spines (without synapsin, as a % of the total number of spines; ***p = 0.0004), but not (B3) unpaired/nonsynaptic spines with synapsin close (ns, not significant; p = 0.61). Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): *** p < 0.001; ns, not significant.

Figure 5. Phosphatase inhibitors prevent NMDA-induced synapse pruning

Cultured neurons were treated twice with NMDA (20 μM for 3 min) alone, NMDA plus a serine-threonine phosphatase inhibitor (CA, fos, OA, or FK506), inhibitor alone, or vehicle-/mock-treated, and processed for immunocytochemistry after 2 hrs. Quantification of the effects of (A) CA, (B) fos, (C) OA, or (D) FK-506 on NMDA-induced changes in synapse density from neurons stained with GFP and synapsin. Data are averaged means of synapses from at least three biological samples. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): *** p < 0.001; ns, not significant.

Figure 6. Caspase-3 inhibition blocks NMDA-induced synapse elimination

(A) 3-D SIM image of MitoTracker Red-stained mitochondria in a dendrite from a GFP-expressing cultured neuron. Scale bar, 2 μm. (B) Cultured neurons were treated twice with NMDA (20 μM for 3 min) alone, NMDA plus caspase-3 inhibitor, z-DEVD-FMK (DEVD), DEVD alone, or vehicle-treated, and processed for immunocytochemistry after 2 hrs. Quantification of the effect of DEVD on NMDA-induced changes in synapse density from neurons stained with GFP and synapsin. Data are averaged means of synapses from three biological samples. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): *** p < 0.001.

Figure 7. Protein synthesis and gene transcription inhibition prevent NMDA-mediated synapse loss

Cultured neurons were treated twice with NMDA (20 μM for 3min) alone, NMDA plus a protein synthesis or transcription inhibitor, inhibitor alone, or vehicle-/mock-treated, and processed for immunocytochemistry after 2 hrs. Quantification of the effects of (A1) anisomycin (aniso), (A2) cycloheximide (CHX), or (B) actinomycin D (actino) on NMDA-induced changes in synapse density from neurons stained with GFP and synapsin. Data are averaged means of synapses from at least three biological samples. Values within bars represent sample sizes (^# neurons). Error bars represent SEM. Significance from control (CTRL): * p < 0.05, *** p < 0.001.