Postsynaptic density-93 deficiency protects cultured cortical neurons from N-methyl-D-aspartate receptor-triggered neurotoxicity

Abstract

It has been reported that N-methyl-D-aspartate receptor (NMDAR)-triggered neurotoxicity is related to excessive Ca(2+) loading and an increase in nitric oxide (NO) concentration. However, the molecular mechanisms that underlie these events are not completely understood. NMDARs and neuronal NO synthase each binds to the scaffolding protein postsynaptic density (PSD)-93 through its PDZ domains. In this study, we determined whether PSD-93 plays a critical role in NMDAR/Ca(2+)/NO-mediated neurotoxicity. We found that the targeted disruption of the PSD-93 gene attenuated the neurotoxicity triggered by NMDAR activation, but not by non-NMDAR activation, in cultured mouse cortical neurons. PSD-93 deficiency reduced the amount of NMDAR subunits NR2A and NR2B in synaptosomal fractions from the cortical neurons and significantly prevented NMDA-stimulated increases in cyclic guanosine 3',5'-monophosphate and Ca(2+) loading in the cortical neurons. These findings indicate that PSD-93 deficiency could block NMDAR-triggered neurotoxicity by disrupting the NMDAR-Ca(2+)-NO signaling pathway and reducing expression of synaptic NR2A and NR2B. Since NMDARs, Ca(2+), and NO play a critical role during the development of brain trauma, seizures, and ischemia, the present work suggests that PSD-93 might contribute to molecular mechanisms of neuronal damage in these brain disorders.

Fig. 1

Efficacy of pCMV-U6-siRNA constructs. (A)RT-PCR analysis showed that a 495 bp product from PSD-93 was inhibited significantly by siP3, GAPDH mRNA was used as a loading Control. siP3 significantly decreased the mRNA of PSD93 by 80.2% of control.(B) Confirmation by western blot analysis of silencing PSD93 expression, The blots were reprobed with anti-GAPDH antibody to verify protein loading. (C) Relative protein level of PSD-93 after siRNA transfection, PSD93 decreased by 79.5% compared to control group when transfected with siP3.

Fig. 2

PSD-93 deficiency or knockdown protects against NMDA-stimulated neurotoxicity. Cortical neurons were cultured from WT and PSD-93 KO mice and treated with NMDA (0, 10, 20, 30, 40, and 60 μM) in the presence of 10 μM CNQX and 2 μM nimodipine. (A) Neuronal viability was assessed by MTT assay. (B) Percentage of neuronal death was determined by propidium iodide and calcein AM staining. * P < 0.05, ** P < 0.01 vs corresponding WT. n = 6 repeats. (C) Representative photographs showing propidium iodide and calcein AM staining of cortical neurons cultured from WT and PSD-93 KO mice treated with or without 30 μM NMDA or 10 μM MK-801+ 30 μM NMDA. Scale bar: 60 μm. (D) acute knock down PSD-93 by siP3 can inhibit the toxicity induced by NMDA at the dose of 30μM. MTT assay shows the cell viability reduced by 44.12% **P<0.01 vs normal group. However, siP3 can alleviate this toxicity, the cell viability increase 17%,# P<0.05 vs group transfected by siLuc.

Fig. 3

MK-801 attenuates NMDA-induced neurotoxicity in cultured cortical neurons from WT but not from PSD-93 KO mice. Percentage of neuronal death was determined by propidium iodide and calcein AM staining (top) and neuronal viability by MTT assay (bottom). * P < 0.05 vs the corresponding NMDA concentration alone. n = 6 repeats.

Fig. 4

PSD-93 deficiency has no effect on non-NMDA receptor-triggered neurotoxicity in cultured cortical neurons. Cultured neurons were treated with kainate at the doses shown in the presence of 10 μM MK-801 and 2 μM nimodipine. Percentage of neuronal death was determined by propidium iodide and calcein AM staining (top) and neuronal viability by MTT assay (bottom). n = 6 repeats.

Fig. 5

PSD-93 deficiency significantly reduces synaptic expression of NR2A and NR2B in cultured cortical neurons. (A) Representative Western blots showing the levels of PSD-93, PSD-95, NR2A, and NR2B in the total soluble and synaptosomal fractions. (B) Statistical summary of the densitometric analysis expressed relative to WT mice after normalization to corresponding β-actin or N-cadherin. ** P < 0.01 vs WT mice. n = 3 repeats.

Fig. 6

PSD-93 deficiency attenuates NMDA-stimulated Ca²⁺ loading in cultured cortical neurons. Cultures were challenged for 5 or 10 min with 30 μM NMDA, 10 μM CNQX, and 2 μM nimodipine in medium containing ⁴⁵CaCl₂. CMP: counts per minute. * P < 0.05 vs corresponding control. n = 6 repeats.

Fig. 7

PSD-93 deficiency decreases NMDA-stimulated increase in cGMP in cultured cortical neurons. (A) NMDA-induced neurotoxicity was NOS-dependent in cultured cortical neurons. Cultured neurons were treated with NMDA (30 μM or 60 μM) with or without L-NAME (10 μM or 30 μM). Neuronal viability was determined by MTT assay (top) and percentage of neuronal death by propidium iodide and calcein AM staining (bottom). * P < 0.05 vs NMDA alone. n = 6 repeats. (B) Cultured neurons from WT and PSD-93 KO mice were exposed to 30 μM NMDA. PSD-93 deficiency decreased the NMDA-stimulated increase in cGMP level in cultured cortical neurons. * P < 0.05 vs the corresponding NMDA-treated group from WT mice. n = 6 repeats.