Fig 2: GluA1 CTD is not essential for spatial learning and memory.(A) Learning curves showing the distance covered to find a hidden platform in the Morris water maze per training day (average of 4 trials/day) in WT (open circles) and HA-?CTD GluA1 KI (filled circles) mice. Mixed effects analysis revealed that the distance necessary to find the platform decreased during training in both groups (day effect, p<0.0001). Although there was only a trend toward a genotype effect (p=0.0539), there was a significant interaction between day and genotype (p<0.05). Distance covered to find a cued platform across 2 days (C1 and C2) is shown in the right side of the plot and showed a significant effect of day (p<0.01) but not genotype (p=0.259), and there was no significant day x genotype interaction (p=0.511). n = 12 WT, 15 KI. (B) Probe trial results showing the number of crossings over the location under which the platform was hidden in the target quadrant during training (circles, empty bars) and over equivalent positions in non-target quadrants (squares, patterned bars) in a 60-s trial performed the day after the last training session. WT mice are represented by empty shapes and HA-?CTD GluA1 KI by filled shapes. n = 12 WT, 14 KI. Both genotypes showed a clear preference for the target location vs non-target locations (WT, p=0.0010, KI, p=0.0012,). WT and KI mice did not differ significantly in how many times they crossed the target location (p=0.582 by Mann-Whitney U test). **p<0.01; ***p<0.001 by Wilcoxon paired t-test. Individual mouse values and mean ± SEM are indicated.Figure 3—source data 1.Contains source data for Figure 3.GluA1 CTD is not essential for spatial learning and memory.

Fig 3: Differential redistribution of synaptic a3-NKA, GluA2-AMPA, and GluN2B-NMDA in primary neurons exposed to distinct a-Syn fibrillar polymorphs. (A–J) Immunocytochemistry on DIV 21 of (A) a3-NKA (red) and Homer (green), (C) GluA2-AMPA receptors (red) and PSD95 (green), (E) GluN2B-NMDA receptors (red) and Homer (green) (G), GluA1-AMPA receptors (red) and PSD95 (green), and (I) mGluR5 receptors (red) and Homer (green). Primary neurons were exposed (15 min, 250 nM) to the fibrillar a-Syn polymorphs fibrils, ribbons, and fibrils-91 on DIV 14. Quantification of intensity of synaptic a3-NKA (B), GluA2-AMPA receptors (D), GluN2B-NMDA receptors (F), GluA1-AMPA receptors (H), and mGluR5 (J) clusters (indicative of size; see Material and Methods) was performed and plotted. The box plot shows median, interquartile range, and 10–90% distribution. Number of images (n) analyzed and plotted: (B) 47 from four independent experiments, (D) 37 from three independent experiments, (F) 30 from three independent experiments, (H) 37 from three independent experiments, and (J) 20 from two independent experiments. A Mann-Whitney test was performed. *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. Dot plot shows the averaged value per experiment. Scale bars, 5 µM.

Fig 4: Multiplicative scaling of excitatory synapses using rank order analysis. A–C, Mice hippocampal neurons of DIV14 treated for 48 h with 2 µM TTX to study homeostatic synaptic scaling. The cultures were fixed and immunolabelled for the C terminus of AMPA receptor subunits GluA1 (A), GluA2 (B), and a PSD protein PSD95 (C). D–F, H–J, Representative dendritic compartments of the images in A–C with PSD95 in magenta and GluA1 (D–F) or GluA2 (H–J) in green. Scale bars: 10 µm (A–C) and 5 µm (D–F, H–J). G, K, A histogram was plotted for the average intensity of GluA1 (G) or GluA2 (K) per PSD95 puncta between the control and the TTX-treated conditions (n = 534 and 454 synapses for control and TTX dataset, respectively, p < 10-6 for GluA1 and p < 10-3 for GluA2, respectively, KS test). L, O, A random sample of 400 intensity values were chosen, rank ordered and plotted to provide a scaling equation for GluA1 (L; y = 1.56x – 6890) and GluA2 (O; y = 1.22x - 1190). M, P, Using the slope of the equation as a reference, the TTX dataset was scaled using multiple values and compared with the control. The scaling factor providing the maximum p value between the scaled-TTX and control dataset was chosen as the multiplicative scaling factor for GluA1 (1.28, p = 0.006, KS test; M) and GluA2 (1.16, p = 0.80, KS test; P). N, Q, Cumulative frequency distribution between control, TTX, and scaled-TTX for GluA1 (N) and GluA2 (Q) showed no significant difference between the control and scaled-TTX distributions.

Fig 5: GluA1/A2CTD supports LTP.(A) Left panel, schematic illustration of the recombinant AMPAR subunits employed to replace endogenous AMPAR: GluA1/A2CTD and (edited) GluA2 (R) in hippocampal CA1 pyramidal neurons from Gria1-3 f/f mice. Note that these two subunits form heteromeric, non-rectifying AMPAR (see Figure 1—figure supplement 2). ATD, amino-terminal domain; LBD, ligand-binding domain; TM, transmembrane domain; CTD, carboxy-terminal domain. Middle panel, summary and timeline of the experiment. Right panel, schematic illustration of the experimental setup with simultaneous whole-cell recordings from control and transfected CA1 pyramidal neurons. (B) Scatterplot measuring the baseline EPSC size at -70 mV in control (X axis) and Cre + GluA1/A2CTD + GluA2 (R) expressing (Y axis) neurons. Open circles represent individual pairs of control and transfected neurons, filled circle represents mean ± SEM. Inset shows sample traces from a control (black trace) and a transfected (green trace) cell. n = 16 pairs. p=0.804, two-tailed Wilcoxon signed-rank test. (C) Plot representing the mean ± SEM EPSC at -70 before and after LTP induction (arrow) normalized by the average baseline EPSC size (dashed gray line) in control (filled circles) and Cre + GluA1/A2CTD + GluA2 (R) expressing (green circles) CA1 pyramidal neurons. Sample traces before and 45’ after LTP induction in control (black traces) and transfected (green traces) CA1 pyramidal neurons are shown to the right of the plot. n initial/final = 13/7 control, 17/10 transfected neurons. p=0.775 (min. 45), unpaired t-test. (D) Left panel, schematic illustration of the recombinant AMPAR subunit employed to replace endogenous GluA1: GluA1/A2CTD in hippocampal CA1 pyramidal neurons from Cas9 KI mice. Middle panel, summary and timeline of the experiment. Right panel, schematic illustration of the experimental setup with simultaneous whole-cell recordings from control and transfected CA1 pyramidal neurons. (E) Scatterplot measuring the baseline EPSC size at -70 mV in control (X axis) and Gria1 gRNA + GluA1/A2CTD expressing (Y axis) neurons. Open circles represent individual pairs of control and transfected neurons, filled circle represents mean ± SEM. Inset shows sample traces from a control (black trace) and a transfected (green trace) cell. n = 12 pairs. p=0.557, two-tailed Wilcoxon signed-rank test. (F) Plot representing the mean ± SEM EPSC at -70 mV before and after LTP induction (arrow) normalized by the average baseline EPSC size (dashed gray line) in control (filled circles) and Gria1 gRNA + GluA1/A2CTD expressing (green circles) CA1 pyramidal neurons. Sample traces before and 45’ after LTP induction in control (black traces) and transfected (green traces) CA1 pyramidal neurons are shown to the right of the plot. Scale bars: 50 pA, 50 ms. n initial/final = 8/8 control, 11/9 transfected neurons. p=0.683 (min 45), unpaired t-test.Figure 1—source data 1.Contains source data for Figure 1.GluA1/A2CTD supports LTP.