Shank3 modulates Rpl3 expression and protein synthesis via mGlu5: implications for Phelan McDermid syndrome

Abstract

Mutations or deletions in the SHANK3 gene have been identified in up to 1% of autism spectrum disorder cases and are considered the primary cause of neuropsychiatric symptoms in Phelan McDermid syndrome (PMS). While synaptic dysfunctions have been extensively documented in the absence of Shank3, other mechanisms through which Shank3 may regulate neuronal functions remain unclear. In this study, we report that the ribosomal protein Rpl3 and overall protein synthesis are downregulated in the cortex and striatum of Shank3 knockout (KO) mice and in neurons differentiated from human-induced pluripotent stem cells (hiPSCs) derived from a PMS patient. Moreover, restoring Rpl3 expression in the striatum of Shank3 KO mice was sufficient to rescue protein synthesis and mitigate excessive grooming, suggesting that the behavioral alterations observed in Shank3 KO mice might be, at least in part, caused by Rpl3 downregulation and consequent impaired protein synthesis. Furthermore, we demonstrated that chronic inhibition of mGlu5 is sufficient to reduce Rpl3 expression, which in turn impairs global protein synthesis. Consequently, chronic treatment with VU0409551, a potent and selective mGlu5 positive allosteric modulator, rescues Rpl3 expression and the resulting reduction in protein synthesis, leading to long-lasting improvements in behavioral deficits in Shank3 KO mice Altogether, we propose a new role for Shank3 in modulating Rpl3 protein expression, ribosomal function, and protein synthesis by downregulating mGlu5 receptor activity.

Fig. 1. Rpl3 transcript and protein are down regulated in the cortex and striatum of Shank3 KO mice.

a Hierarchical biclustering (heatmap) of all the RNAseq genes in the cortex and striatum of WT (n = 3) and Shank3 KO (n = 3) mice. The colour spectrum from green to red represents the gene expression intensity from low to high, respectively. b Venn diagram shows upregulated DEGs in statistical comparisons contrasting genotypes (WT vs KO, shaded red) and regions (the cortex, CTX and striatum, STX, shaded blue) and their interactions (shaded green). The numbers indicate the number of DEG genes in each group. c Similar Venn diagrams for downregulated DEGs. d Logfold changes of Rpl3 shows a lower expression in the cortex and striatum of Shank3 KO mice. e and f Representative blots of Rpl3 expression e and relative quantification f of Rpl3 expression in total lysate and synaptosomes fraction (syn) from cortex of WT and Shank3 KO mice. Rpl3 protein levels were normalized to the corresponding βIII-Tubulin, and ratios were compared between genotypes. Cortex total lysate n = 6 per genotype; cortex synaptosomes fraction n = 4 per genotype. g and h Representative blots of Rpl3 expression g and relative quantification h of Rpl3 expression in total lysate and synaptosomes fraction (syn) from striatum of WT and Shank3 KO mice. Rpl3 protein levels were normalized to the corresponding βIII-Tubulin, and ratios were compared between genotypes. Striatum total lysate n = 6 per genotype; Striatum synaptosomes fraction n = 4 per genotype. i Comparison of Rpl3 signal imaged by both confocal and STED microscopy, scale bar 3μm. j and k Rpl3 signal resolved with STED microscopy show a reduction in area covered by signal and number of puncta in both cortex j and striatum k of Shank3 KO mice when compared with WT mice. k Shank3 KO mice show a reduction in striatum in mean intensity of Rpl3 signal when compared with WT mice. Cortex and striatum: n = 3 per genotype. All data are shown as box and whiskers plot. All Data are presented as Min to Max; all p-values were derived using the Mann-Whitney test (h total lysate) or the Unpaired t-test (f, h synaptosomes fraction, j and k). *p < 0.05; **p < 0.01; *** p < 0.001.

Fig. 2. Reduced Rpl3 expression impairs protein synthesis in the cortex and striatum of Shank3 KO mice.

a–d Protein synthesis evaluated using the SUnSET assay in both total lysate a and c and synaptosomes fraction b and d of both cortex a and b and striatum c and d. The puromycin signals were normalized to the respective Coomassie staining signals. Cortex total lysate n = 10 per genotype; striatum total lysate n = 9 per genotype; cortex and striatum synaptosomes fraction n = 4. e and f Rpl3 protein levels after infection with AAV9-RPL3 or control in striatum was evaluated in striatum e and cortex f. Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between groups. Cortex and striatum n = 6 per group. g Effect on protein synthesis in Shank3 KO mice after overexpression of AAV9-RPL3. The puromycin signals were normalized to the respective Coomassie staining signals. N = 7 per group. h Self-grooming time in WT and Shank3 KO mice after overexpression of AAV9-RPL3 or control in striatum. WT n = 7 per control, n = 9 per AAV9-RPL3; Shank3 KO n = 7 per control, n = 6 per AAV9-RPL3. All data are shown as box and whiskers plot. All data are presented as Min to Max; all p-values were derived using the Unpaired t-test a–d or the Two-Way ANOVA followed by Holm-Šídák’s multiple comparisons test e–h. *p < 0.05; **p < 0.01; *** p < 0.001.

Fig. 3. Chronic treatment with MTEP, a selective mGlu5 antagonist, decreases Rpl3 expression and protein synthesis.

a Schematic representation of MTEP chronic administration in rat cortical neurons. b Rpl3 protein levels were analyzed by western blot in total lysate from rat cortical neurons at DIV18 after vehicle, 30 min acute treatment or 5 days chronic treatment with MTEP (30 nM). Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between groups. Vehicle and 5 days treatment n = 6, 30 min treatment n = 5. c Protein synthesis levels in rat cortical neurons treat with vehicle, 30 min or 5 days treatment with MTEP (30 nM). The puromycin signals were normalized to the respective Coomassie staining signals. N = 4 per group. d Schematic representation of MTEP (3 mg/kg) administration in WT mice from 7 post-natal day to 14 post-natal days. e and f Protein levels of Rpl3 were analysed by western blot in total lysate obtained from cortex e and striatum f of WT mice treat with MTEP (3 mg/Kg) or vehicle from P7 to P14. Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between groups. Cortex and striatum n = 5 per group. g Schematic representation of MTEP chronic administration in hiPSCs derived neurons. h Cortical neurons derived from iPSCs were treated with either a vehicle or MTEP (30 nM) for 5 days and Rpl3 protein levels was analyzed. The Rpl3 level was normalized to the respective βIII-Tubulin and ratios were compared between groups. N = 4 per group. i Protein levels of Shank3, mGluR5, Rpl3 and Homer1b/c in cortical neuron differentiated from iPSCs derived from a PMS patient and controls. Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between groups. Shank3: n = 7 per group; mGluR5: n = 5 per group; Rpl3: n = 4 per group; Homer1b/c: n = 4 per group. i and j Cortical neurons derived from iPSCs from control and PMS patient were treated with either a vehicle or VU0409551 (40 mM) for 5 days. Rpl3 levels and protein synthesis were analyzed. j The Rpl3 level was normalized to the respective βIII-Tubulin and ratios were compared between groups. Control and PMS n = 5 per group. k The puromycin signals were normalized to the respective Coomassie staining signals. Control and PMS n = 6 per group. All data are shown as box and whiskers plot. All data are presented as Min to Max; all p-values were derived using the One-way ANOVA followed by Tukey’s multiple comparisons test b and c, the Mann-Whitney test (j Rpl3), the Unpaired T-test f–i or the Two-way ANOVA followed by Holm-Šídák’s multiple comparisons test j and K; *p < 0.05; **p < 0.01; *** p < 0.001.

Fig. 4. Fifteen days of chronic treatment with VU0409551 rescues Rpl3 expression and protein synthesis in adult mice (P75).

a Schematic representation of VU0409551 chronic administration in adult mice. b–e Protein levels of Rpl3 were analyzed by western blot in total lysate b and d and synaptosomes fraction c and e obtained from cortex b and c and striatum d and e of WT and Shank3 KO adult mice (post-natal-day 75) after chronic treatment with VU0409551 (5 mg/kg) or vehicle. Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between groups. Cortex total lysate and synaptosomes fraction n = 4 per group; striatum total lysate and synaptosomes fraction n = 4 per group. f and i Protein synthesis level in adult (post-natal day 75) WT and Shank3 KO mice after chronic administration of VU0409551 (5 mg/kg) or vehicle in cortex total lysate (f and Supplementary Fig. 6a), cortex synaptosomes fraction (g and Supplementary Fig. 6b), striatum total lysate (h and Supplementary Fig 6c) and striatum synaptosomes fraction (i and Supplemenatry Fig. 6d). The puromycin signals were normalized to the respective Coomassie staining signals (Supplementary Fig. 6a–d). Cortex total lysate n = 9 per group; cortex synaptosomes fraction n = 4 per group; striatum total lysate n = 6 per group; striatum synaptosomes n = 4 per group. All data are shown as box and whiskers plot. All data are presented as Min to Max; all p-values were derived and calculated using the Two-way ANOVA followed by Holm-Šídák’s multiple comparisons test; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 5. Fifteen days of chronic treatment with VU0409551 rescues Rpl3 expression, protein synthesis and behaviural deficit in young mice (P30).

a and b Protein levels of Rpl3 were analyzed by western blot in total lysate obtained from cortex and striatum of WT and Shank3 KO young mice (post-natal-day 15). Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between genotypes. Cortex n = 5 per genotype and striatum n = 4 per genotype. c and d Protein synthesis was evaluated in WT and Shank3 KO young mice (post-natal day 15) by the SUnSET assay. Representative images show the level of puromycin incorporated into cortex (c and Supplementary Fig. 7a) and striatum (d and Supplementary Fig. 7b). The puromycin signals were normalized to the respective Coomassie staining signals (Supplementary Fig. 7a and Supplementary Fig. 7b). Cortex n = 6 per genotype and striatum n = 8 per genotype. e Schematic representation of VU0409551’s chronic administration in young mice (P15-P30). f and g Protein levels of Rpl3 were analysed by western blot in total lysate obtained from cortex f and striatum g of WT and Shank3 KO young mice (post-natal day 30) after chronic treatment with VU0409551 (5 mg/kg) or vehicle. Protein levels were normalized to the respective βIII-Tubulin and ratios were compared between genotypes. Cortex and striatum n = 7 per group. h and i Protein synthesis level in young (post-natal day 30) WT and Shank3 KO mice after chronic administration of VU0409551 (5 mg/kg) or vehicle in cortex (h and Supplementary Fig. 7d) and striatum (i and Supplementary Fig. 7e) using SUnSET assay. The puromycin signals were normalized to the respective Coomassie staining signals (Supplementary Fig. 7d and Supplementary Fig. 7e). Cortex and striatum n = 6 per group. j Schematic representation of chronic VU0409551 (5 mg/kg) or vehicle administration in young mice (P15-P30), as well as a timetable for behavioral experiments. k Discrimination index derived from the novel object recognition test. WT n = 3 per group; Shank3 KO n = 3 per vehicle, n = 6 per VU0409551. l Self-grooming behavior was evaluated as time spent grooming at different time point. WT vehicle n = 12, KO vehicle n = 6, WT VU0409551 n = 11, Shank3 KO VU0409551 n = 10. m Self-grooming behaviour was evaluated as time spent grooming in animals treated with vehicle or VU0409551 (5 mg/Kg) in WT and Shank3 KO mice. WT n = 7 per group; Shank3 KO n = 3 per vehicle and n = 7 per VU040955. All data are shown as box and whiskers plot b–k and m or Mean and Error l. All data are presented as Min to Max b–k and m or as the mean ± SEM l; all p-values were derived using the Unpaired T-test b and d, the Mann-Whitney test c or the Two-way ANOVA followed by Holm-Šídák’s multiple comparisons test f–i; *p < 0.05; **p < 0.01; *** p < 0.001. (k-m) all p-values were calculated using the Two-way ANOVA with the Bonferroni post hoc test. *<p 0.05; **p < 0.01; ***p < 0.001 compared to the WT vehicle; $p < 0.05; $$$p < 0.001 when compared to WT VU0409551.

Fig. 6. Acute treatment with VU0409551 rescues behavioral defects observed in Shank3 KO mice.

a and b Social behaviors are scored in sociability a and social novelty b tests. WT n = 7 per group; Shank3 KO n = 7 per vehicle n = 6 per VU040955. c Discrimination index evaluated in the novel object recognition test after acute treatment with vehicle or VU0409551 (5 mg/Kg) in WT and Shank3 KO mice. WT n = 7 per group; Shank3 KO n = 7 per vehicle n = 6 per VU0409551. d Social dominance expressed as a percentage of won matches in a tube test after acute treatment with vehicle or VU0409551 (5 mg/Kg) in WT and Shank3 KO mice. WT n = 10 per vehicle n = 15 per VU0409551; Shank3 KO n = 10 per vehicle n = 15 per VU0409551. e and g Acquisition and f and h reversal tasks in the T-Maze task were performed after daily treatments; % of animals reaching the criterion e and f and the number of animals reaching the criteria g and h were analyzed. WT n = 12; Shank3 KO n = 8 per vehicle n = 10 per VU0409551. i and j c-Fos levels were evaluated by immunostaining in cortex i and striatum j of WT and Shank3 KO mice treat with VU0409551 or vehicle. Corte and striatum n = 7 per group. All data are shown as box and whiskers plot a–c and g–j, bar diagram d or XY table points and connecting line e and f. All data are presented as Min to Max a–c and g–j); all p-values were derived using the Two-way ANOVA followed by Holm-Šídák’s multiple comparisons test a and b and g–j, the Fisher’s exact test d; *p < 0.05;**p < 0.01;***p < 0.001; ****p < 0.0001 or using the Two-way ANOVA Bonferroni post hoc test c and e, f; *p < 0.05; **p < 0.01; *** p < 0.001 versus the WT vehicle; $$$p < 0.001 when compared to corresponding WT VU0409551.