Fig 2: Glutamate induces synaptic loss prior to neuronal death in a P2Y1R-dependent manner.a Rat hippocampal neurons (DIV14) exposed to glutamate (100 µM; 30 min) display a decrease in the density of the synaptophysin puncta per dendritic length after 12 h, an effect not observed in the presence of MRS2179 (10 µM), as depicted in the representative confocal images (scale bar, 10 µm). Data are mean ± SEM of synaptophysin puncta per dendritic length quantified from four different cultures analyzing a minimum of 2000 µm of MAP2+ structures from two independent coverslips per condition, per culture. b Representative traces of mEPSC recorded in rat hippocampal neurons (DIV14) 12–14 h after the exposure to glutamate (100 µM; 30 min) either in the absence or in the presence of MRS2179 (10 µM). c, d Cumulative histograms (left) and pooled data (right) of the average of c instantaneous frequency or d peak amplitude of mEPSCs. Pooled data are mean ± SEM of the average frequency/amplitude of mEPSCs calculated from 180 s bins recorded from 11 to 15 neurons per condition. *p < 0.05 and **p < 0.01 Kruskal–Wallis with Dunn’s test for frequency and one-way ANOVA with Sidak’s test for amplitude. e Western blot analysis of general synaptic markers, synaptophysin and syntaxin1A, f glutamatergic synaptic markers, vGluT1 and PSD-95 and g GABAergic synaptic markers, gephyrin and vGAT, in total protein extracts prepared from hippocampi of rats previously injected i.p. with saline or kainate (KA,10 mg/kg). Hippocampal protein extracts of the animals injected with KA displayed a significant decrease in the immunoreactivity of all the synaptic proteins evaluated, except for gephyrin (p = 0.1205), in comparison with saline-injected mice, as depicted in the representative Western blots (in left) and summarized in the histograms (in right). The i.c.v. injection of 1 nmol of MRS2500 15 min prior to the administration of KA prevented or attenuated the decrease in the immunoreactivity of the different synaptic markers. MRS2500 per se did not cause any significant modification in the density of the synaptic markers. Data are percentage (mean ± SEM) of the ratio of immunoreactivity synaptic markers/loading control relative to the average ratio obtained in samples from aCSF/Sal group (n = 4). *p < 0.05, **p < 0.01, and ***p < 0.001 one-way ANOVA with Sidak’s post hoc test for KA vs. respective saline

Fig 3: Reduced presynaptic type III Nrg1 signaling impairs glutamatergic synaptic transmission at vHipp-nAcc synapses. A, Schematic of genotype-specific in vitro circuits. Glutamatergic transmission at vHipp-nAcc synapses was examined in gene chimeric cocultures. Experiments show ventral hippocampus/subiculum slices from an individual Nrg+/+ or Nrg+/− mouse plated as micro thinned explants. Dispersed nucleus accumbens neurons from WT mice were added 24 h later. B, Representative traces of spontaneous glutamate-receptor mediated synaptic activity (Glu mEPSCs: bicuculline and TTX resistant; CNQX-APV sensitive) recorded from +/+ vHipp to +/+ nAcc synapses (top) and from +/− vHipp to +/+ nAcc synapses (bottom). C, Box plots of mEPSC frequency data from +/+ vHipp to +/+ nAcc and from +/− vHipp to +/+ nAcc reveal more than a 3-fold difference (**p < 0.01, n = 8 for each condition) at gene chimeric synapses compared with WT- WT synapses. D, Box plots of mEPSC amplitude data from +/+ vHipp to +/+ nAcc and from +/− vHipp to +/+ nAcc reveal a 2-fold difference (**p < 0.01, n = 8 for each condition) at gene chimeric synapses compared with WT- WT synapses. E, Examination of pre and postsynaptic markers in gene chimeric vHipp-nAcc cocultures. After 5-7 d in vitro, cocultures were fixed, permeabilized, and stained with antibodies targeted to vesicular glutamate transporter 1 (vGluT1; red) and to PSD95 (green). Red “clusters” of vGluT1 are colocalized with PSD95 (green) on neurites of dispersed nAcc neurons innervated by +/+ vHipp (E, top left) or +/− vHipp (E, bottom left; scale bar, 10μm). The arrows indicate colocalization of PSD95 with vGLuT1 (yellow puncta) along neurites of nAcc MSNs. The number of PSD95 positive/vGluT1 clusters along neurites of dispersed nAcc neurons innervated by +/+ vHipp were significantly greater than those innervated by +/− vHipp inputs (28 ± 4 per 100 μm, n = 9, 5 vs 17 ± 3 per 100 μm n = 6, 3; where n= the number of samples/experiment and the number of separate experiments; **p < 0.01; E, middle, left). The number of PSD95 clusters along neurites of nAcc neurons were counted and the bar graph (E, middle, right) showed no difference between +/+ and +/− vHipp innervation (+/+vHipp to +/+nAcc: 60 ± 10, n = 9, 5 vs +/−vHipp to +/+nAcc: 54 ± 10, n = 6, 3; p > 0.05). The number of vGluT1 clusters along neurites of nAcc neurons were also counted and the bar graph (E, right) showed no difference between +/+ and +/− vHipp innervation (+/+ vHipp to +/+ nAcc: 31 ± 4, n = 9, 5 vs +/− vHipp to +/+ nAcc: 29 ± 8, n = 6, 3; p > 0.05). F, Examination of surface versus total glutamate A2/A3 receptor subtypes (GluA2) in gene chimeric vHipp-nAcc cocultures. After 5-7 d in vitro, cocultures were stained with antibodies targeted to GluA2. For labeling of sGluA2 (green), the cultures were incubated with anti-GluR2 antibody, extracellular, for 45 min before fixation, and then, cultures were fixed, permeabilized, and total GluA2 (red) were recognized with anti-GluR2 + GluR3 antibody, C terminal. Clusters of surface (green) vs total GluA2/3 (red) can be found on neurites of dispersed nAcc neurons innervated by vHipp axons (F, left panel; scale bar, 5 μm). The ratio of integrated intensities of sGluA2 versus total GluA2 between gene chimeric conditions differed in a statistically significant manner (+/+ vHipp to +/+ nAcc: 0.36 ± 0.07, n = 10, 4 vs +/− vHipp to +/+nAcc: 0.18 ± 0.02, n = 8, 3; **p < 0.01; F, right-hand panel). G, Examination of synapse formation of sGluA2 (red) and vGluT1 (green) along nAcc neurites in vHipp-nAcc cocultures. After 5-7 d in vitro, cocultures were incubated with anti-GluR2 antibody recognizing an extracellular epitope, for 45 min to label sGluA2, and then, cultures were fixed, permeabilized, and stained with antibody targeted to vGluT1. Representative micrographs of vGluT1 and sGluA2 colocalization are shown, the arrows are examples of those colocalized sites (yellow puncta) of sGluA2 (red) and vGLuT1 (green) along neurites of nAcc MSNs (G; scale bar, 5 μm).

Fig 4: Loss of synaptic ribbons and piccolino from the AZs of RIBEYE-deficient IHCs.(A) Simplified schematic representation of an IHC with the afferent and efferent connectivities. (B) Maximal projection of confocal sections from organs of Corti immunolabeled for CtBP2 and RIBEYE, present in the nuclei and the ribbons, respectively. The RBEWT/WT staining (top row) shows small puncta in the outer hair cell (OHC) and IHC rows representing the synaptic ribbons, which are completely absent in the RBEKO/KO hair cells (bottom row). Scale bar = 50 µm. Zoom into the IHC row (right column), emphasizes the complete disappearance of CtBP2-labeling at the basolateral part of RBEKO/KO IHCs. Scale bar = 10 µm. (C) Maximal projection of confocal sections from organs of Corti co-labeled for the presynaptic marker and anchor of the ribbon, bassoon (left column), and the postsynaptic marker, PSD-95 (middle column), in RBEWT/WT and RBEKO/KO IHCs. The merged picture (right column) shows the juxtaposition of bassoon (magenta) with PSD-95 (green), indicating its presence both at RBEWT/WT and ribbonless RBEKO/KO IHC synapses. Scale bar = 1 µm. (D) Maximal projection of confocal sections from organs of Corti co-labeled for the presynaptic marker RIM2 (left column) and the postsynaptic marker PSD-95 (middle column). The merged picture (right column) shows the co-localization of RIM2 (magenta) with PSD-95 (green) meaning its presence at the ribbonless IHC pre-synapses (scale bar = 5 µm). (E) Maximal projection of confocal sections from organs of Corti co-labeled for piccolino, a specific short splice variant of piccolo found at ribbons of RBEWT/WT IHC synapses (left column), co-labeled with PSD-95 (middle column). The merged pictures (right column) show PSD-95 (green) immunofluorescence lacking juxtaposed piccolino signal (magenta) in RBEKO/KO (bottom row), indicating absence of piccolino from afferent synapses of mutant IHCs. The punctate labeling for piccolo, away from PSD-95, most likely represents labeling of piccolo at conventional efferent synapses (schematically shown in Figure 1A). Scale bar = 5 µm. (F) Quantification of the area of PSD-95 immunofluorescent spots. The PSD-95 spots are siginificantly bigger in the RBEKO/KO IHCs (p<0.0001, Mann-Whitney-Wilcoxon test, n = 178 spots, N = 3 for RBEKO/KO and n = 163 spots, N = 3 for RBEWT/WT). Box plots show 10, 25, 50, 75 and 90th percentiles with individual data points overlaid; means are shown as crosses.

Fig 5: Results from the gene expression study in the hippocampus with GAPDH used as the reference gene. As shown in (E), expression of DLG4 (the gene for PSD-95) appeared to be increased in ?4 carriers.