Fig. 2

Genetically encoding GluA2(R) in J20 mice rescues neuronal and spine numbers but does not alter Aβ-pathology. a NeuN⁺ neurons in the hippocampus of WT, GluA2^G/G, J20 and GluA2^G/G/J20 mice. Stereological quantification of neuronal numbers revealed a significant decrease in the CA1 neuronal population of J20 mice compared to all other genotypes (n = 5/genotype) while no changes to CA3 neuronal populations were observed (n = 5/genotype). b Representative images of Golgi-stained CA1 dendritic spines in WT, GluA2^G/G, J20 and GluA2^G/G/J20 mice. Quantification of spine density revealed a significant reduction of spines in J20 mice compared to all other genotypes (n = 3 dendritic branches/neuron with 3 neurons/brain and 5 brains/genotype). Both GluA2^G/G and GluA2^G/G/J20 mice displayed significantly higher spine density than WT littermates. c Scholl analysis and representative traces of Golgi-impregnated neurons from the hippocampal CA1 revealed significantly reduced numbers of dendritic intersections in J20 mice at varying distances from the soma compared to WT (*), GluA2^G/G (^) and GluA2^G/G/J20 (#) animals (n = 3 neurons/brain and 5 brains/genotype). d Representative image of Thioflavin-S⁺ plaques in J20 and GluA2^G/G/J20 mice. Quantification revealed no significant differences in Thioflavin-S⁺ plaques (n = 8/genotype). No differences in other Aβ species including (e) total 6E10⁺ area (n’s; J20 = 5, GluA2^G/G/J20 = 4), (f) Aβ40 (n = 8/genotype), (g) Aβ42 (n = 13/genotype), or (h) total soluble and insoluble Aβ (n = 5/genotype) were observed between J20 and GluA2^G/G/J20 mice. a-b Analysed using univariate ANOVA with Bonferroni correction, c analysed using two-way ANOVA with Bonferroni correction, d-g analysed using t-tests with Welch’s corrections and h analysed using either Mann-Whitney test (soluble fraction) or unpaired t-test (insoluble fraction). Each value represents the mean ± the SD for bar graphs and SEM for line graphs. **p < 0.01, ***p < 0.001, ****p < 0.0001