Figure 4. Loss of parkin function increases surface GluK2 levels, KAR currents and causes excitotoxicity.

(a) Surface biotinylation assay in hippocampal neurons infected with lentivirus encoding sh-scrambled or sh-parkin, or sh-parkin+parkin^R. Endogenous parkin silencing increased surface GluK2 levels; co-infection with lentivirus encoding parkin^R rescued GluK2 surface levels (one-way analysis of variance (ANOVA)–Dunnett’s multiple comparison test, **P=0.0053, F=14.25). (b) Surface Myc labelling in hippocampal neurons co-transfected with sh-parkin+Myc-GluK2a or sh-scrambled+Myc-GluK2a, or sh-parkin+parkin^R+Myc-GluK2a (one-way ANOVA–Bonferroni test, **P=0.0066, F=6.082, 29 degrees of freedom; ten neurons analysed/each condition). Scale bar, 20 μm. (c) KAR current analysis in hippocampal neurons infected with lentivirus encoding GFP bicistronic sh-scrambled or sh-parkin, or sh-parkin+parkin^R. Whole-cell responses were induced by rapid application of 100 μM kainate. To isolate KAR currents from those resulting from AMPAR opening, we added 10 μM GYKI 53655 (IC₅₀=0.9±0.08 μM)16. Responses are shown before application (left panel), during the concomitant application of GYKI53655 (central panel) and after GYKI53655 washout (right panel). Western blotting shows parkin expression in the three experimental conditions. The chart plot shows the ratio between current stimulated by kainate+GYKI53655 and the current triggered by 3 s application of kainate alone (one-way ANOVA and Tukey test, **P<0.01; ***P<0.001; F=15.87). Error bars indicate 25th and 75th percentiles. (d) Primary hippocampal neurons were infected with lentivirus encoding GFP bicistronic sh-scrambled or sh-parkin, or sh-parkin+parkin^R. Data represent the percentage of infected cells (green fluorescent cells) labelled by propidium iodide (PI). Kainate 2–4 μM or concanavalin A 200 μg ml⁻¹ alone did not induce cell death. Kainate 2–4 μM+concanavalin A 200 μg ml⁻¹ caused excitotoxicity in parkin-silenced cells. Co-infection with lentivirus encoding parkin^R rescued cell death (one-way ANOVA–Tukey test, *P<0.05 versus sh-scrambled 2 μM kainate; ***P<0.001 versus sh-scrambled 4 μM kainate; F=20.02). NS102 20–150 μM blocked excitotoxicity in parkin-silenced neurons (one-way ANOVA–Dunnett’s multiple comparison test, *P<0.05, **P<0.01, versus kainate treated; ^###P<0.001 versus untreated, F=7.268). GYKI53655 (10 μM) did not rescue cell death. (e) Western blottings for cleaved spectrin and cleaved calcineurin A in substantia nigra lysates from 3-week-old controls (n=7) and littermate parkin-Q311X mice (n=6). Results derive from three independent experiments (unpaired t-test, **P=0.0096, t=3.128 for cleaved spectrin; *P=0.0124, t=2.941 for cleaved calcineurin A). All error bars in Fig. 4 histograms indicate±s.e.m.