Ischemic insults direct glutamate receptor subunit 2-lacking AMPA receptors to synaptic sites

Abstract

Regulated AMPA receptor (AMPAR) trafficking at excitatory synapses is a mechanism critical to activity-dependent alterations in synaptic efficacy. The role of regulated AMPAR trafficking in insult-induced synaptic remodeling and/or cell death is, however, as yet unclear. Here we show that brief oxygen-glucose deprivation (OGD), an in vitro model of brain ischemia, promotes redistribution of AMPARs at synapses of hippocampal neurons, leading to a switch in AMPAR subunit composition. Ischemic insults promote internalization of glutamate receptor subunit 2 (GluR2)-containing AMPARs from synaptic sites via clathrin-dependent endocytosis and facilitate delivery of GluR2-lacking AMPARs to synaptic sites via soluble N-ethylmaleimide-sensitive factor attachment protein receptor-dependent exocytosis, evident at early times after insult. The OGD-induced switch in receptor subunit composition requires PKC activation, dissociation of GluR2 from AMPA receptor-binding protein, and association with protein interacting with C kinase-1. We further show that AMPARs at synapses of insulted neurons exhibit functional properties of GluR2-lacking AMPARs. AMPAR-mediated miniature EPSCs exhibit increased amplitudes and enhanced sensitivity to subunit-specific blockers of GluR2-lacking AMPARs, evident at 24 h after ischemia. The OGD-induced alterations in synaptic AMPA currents require clathrin-mediated receptor endocytosis and PKC activation. Thus, ischemic insults promote targeting of GluR2-lacking AMPARs to synapses of hippocampal neurons, mechanisms that may be relevant to ischemia-induced synaptic remodeling and/or neuronal death.

Figure 1.

A, Time course of OGD-induced hippocampal neuronal death. Representative images of PI uptake and staining for the neuronal marker neuronal-specific nuclear protein (NeuN) showing OGD-induced delayed neuronal death in hippocampal cultures at 0, 24, and 48 h. Neurons at 14–21 DIV were subjected to 20 min OGD. B, Summary data of PI uptake. OGD induced neuronal death, evident at 24 and 48 h after insult (*p < 0.05, difference from sham groups; ^#p < 0.05, difference from 0 h; sham, 0 h, n = 236 neurons; 24 h, n = 217; 48 h, n = 198; OGD, 0 h, n = 241; 24 h, n = 205; 48 h, n = 227). Scale bar, 50 μm.

Figure 2.

OGD decreases GluR2 surface expression at synaptic sites in cultured hippocampal neurons. A, Representative images showing the juxtaposition of surface GluR2 and the presynaptic marker synaptophysin in control and experimental neurons at 24 h after OGD. Higher-magnification images correspond to the boxed areas in the lower-magnification images. Scale bars: lower-magnification images, 10 μm; higher-magnification images, 5 μm. B, OGD does not significantly alter total GluR2 protein expression at 0, 12, and 24 h (n = 6 for each group; p > 0.05). C, OGD reduces surface GluR2 expression, assessed at 0, 12, and 24 h after insults (n = 6 for each group; *p < 0.05). D, Quantification of the percentage of synaptophysin-positive synapses containing GluR2 puncta in control and OGD neurons (n = 33, 29, 32, and 26, respectively, per group; *p < 0.05). E, Quantification of the percentage of surface GluR2 puncta that are juxtaposed with synaptophysin puncta in control and OGD neurons (n = 29, 30, 31, and 26, respectively, per group; *p < 0.05).

Figure 3.

OGD does not significantly alter GluR1 and GluR3 surface expression at synaptic sites. A, Representative images showing juxtaposition of GluR1 and GluR3 puncta with synaptophysin puncta in control and experimental hippocampal neurons in culture at 24 h after OGD. Higher-magnification images are of boxed areas indicated in the lower-magnification images. Scale bars: lower-magnification images, 15 μm; higher-magnification images, 5 μm. B, F, OGD (20 min) does not significantly alter total GluR1 (B) or GluR3 (F) protein expression assessed at 0, 12, and 24 h after OGD (B, n = 6 for each group, p > 0.05; F, n = 6 for each group, p > 0.05). C, G, OGD does not detectably alter GluR1 (C) and GluR3 (G) surface expression, as assessed by colorimetric assay (C, n = 6 for each group, p > 0.05; G, n = 6 for each group, p > 0.05). D, H, Quantification of the percentage of synaptophysin-positive synapses containing GluR1 (D) or GluR3 (H) puncta in control and OGD neurons (D, n = 25, 25, 32, and 28, respectively, per group, p > 0.05; H, n = 30, 29, 30, and 27, respectively, per group, p > 0.05). E, I, Quantification of the percentage of surface GluR1 (E) or GluR3 (I) puncta that are juxtaposed with synaptophysin puncta in control and OGD neurons (E, n = 26, 27, 29, and 24, respectively, per group, p > 0.05; I, n = 26 for each group, p > 0.05).

Figure 4.

OGD-induced retrieval of GluR2 subunits from synaptic sites occurs via clathrin-dependent internalization. A, Representative image showing that hypertonic sucrose (0.45 m, 20 min before OGD) blocks OGD-induced loss of synaptic GluR2 subunits, assessed at 0 h after OGD. Scale bar, 10 μm. B, Pretreatment with hypertonic sucrose does not significantly alter total GluR2 expression, assessed at 0 h after OGD (n = 6 per group; p > 0.05). C, Hypertonic sucrose pretreatment blocks OGD-induced loss of surface GluR2 subunits, assessed at 0 h after OGD (n = 6 per group; *p < 0.05). D, E, Hypertonic sucrose prevents OGD-induced decrease of synaptic GluR2 subunits, assessed at 0 h after OGD (D, n = 23 per group, *p < 0.05; E, n = 26 per group, *p < 0.05). F, Left, OGD promotes association of GluR2 subunits with the AP2 complex. Protein samples from control and OGD neurons were immunoprecipitated with an anti-adaptin β₂ antibody, and resulting samples were sequentially immunoblotted with antibodies to GluR2 and adaptin β₂. Right, Summary data for three independent experiments involving different batches of neurons (*p < 0.05).

Figure 5.

OGD promotes internalization of GluR1 and GluR3 subunits via clathrin-dependent endocytosis pathway. A, F, Representative images showing that hypertonic sucrose promotes OGD-induced increase in expression of GluR1 (A) and GluR3 (F) subunits at synaptic sites, assessed at 0 h after OGD. Scale bars, 10 μm. B, G, Hypertonic sucrose does not significantly alter total GluR1 (B) and GluR3 (G) subunit expression, assessed at 0 h after OGD (B, n = 6 per group, p > 0.05; G, n = 6 per group, p > 0.05). C, H, Hypertonic sucrose promotes OGD-induced increase in surface expression of GluR1 (C) and GluR3 (H) subunits, assessed at 0 h after OGD (C, n = 6 per group, *p < 0.05; H, n = 6 per group, *p < 0.05). D, E, I, J, Hypertonic sucrose promotes OGD-induced increase in GluR1 (D, E) and GluR3 (I, J) subunit expression (D, I) and fractional expression (E, J) at synaptic sites, assessed at 0 h after OGD (D, n = 22 per group; E, n = 26 per group; I, n = 25 per group; J, n = 23 per group.;*p < 0.05).

Figure 6.

OGD enhances delivery of GluR1 and GluR3 occurs via SNARE-dependent exocytosis. A, F, Representative images showing that BoNT A markedly attenuates OGD-induced increase in GluR1 (A) and GluR3 (F) subunits at synaptic sites in the presence of hypertonic sucrose (0.45 m, 20 min) at 0 h after OGD. Scale bar, 10 μm. B, G, BoNT A does not detectably alter total cellular GluR1 (B) and GluR3 (G) expression at 0 h after OGD (B, n = 6 per group, p > 0.05; G, n = 6 per group, p > 0.05). C, H, BoNT A blocks OGD-induced increase of GluR1 (C) and GluR3 (H) surface expression in the presence of hypertonic sucrose at 0 h after OGD (C, n = 6 per group, *p < 0.05; H, n = 6 per group, *p < 0.05). D, E, I, J, BoNT A markedly inhibits OGD-induced increase of GluR1 (D, E) and GluR3 (I, J) synaptic expression in the presence of hypertonic sucrose at 0 h after OGD (D, n = 24 for per group; E, n = 23 per group; I, n = 27 per group; J, n = 25 per group; *p < 0.05).

Figure 7.

OGD enhances the association of GluR2 with PICK1 and reduces its association with ABP in cultured hippocampal neurons. The PKC inhibitor GÖ 6976 blocks the increase in association with PICK1 and the reduction in association with ABP. A, Top, Homogenates prepared from control and OGD hippocampal neurons were precipitated with an anti-GluR2 antibody, and bound protein (PICK1) was detected by immunoblot (IB). Bottom, Summary data for three independent experiments (*p < 0.05). B, Top, Homogenates prepared from control and OGD hippocampal neurons were precipitated with an anti-GluR2 antibody, and bound protein (ABP) was detected by immunoblot (IB). Bottom, Summary data for three independent experiments (*p < 0.05). C, PICK1 or ABP did not coimmunoprecipitate with NR1 subunit of NMDA receptors in hippocampal neurons.

Figure 8.

OGD induces increased expression of functional GluR2-lacking AMPA receptors at postsynaptic membrane. A, Sample traces of AMPA receptor-mediated mEPSCs recorded in control (left) and OGD-treated (right) neurons at 24 h after OGD. B, Quantification data indicate that OGD causes significant increases of AMPA-mEPSC amplitudes that are abolished by PKC inhibitor GÖ 6976 (0.1 μm) at 0, 12, and 24 h after insults (n = 9 per group; p < 0.05). C, OGD has no significant effects on the AMPA-mEPSC frequencies (n = 9 per group; p > 0.05). Data are normalized to the corresponding control values. D, Quantification data indicate that Naspm produces a significantly greater inhibition of the mEPSC amplitudes in OGD versus control neurons, and PKC inhibitor GÖ 6976 prevents the inhibitory effects (n = 10 per group; *p < 0.05). Data are normalized to the corresponding control values. E, Naspm does not significantly alter the mEPSC frequency in either the control or OGD neurons (n = 10 per group; p > 0.05). Data are normalized to the corresponding control values.