Fig. 2
From: Enlarged dendritic spines and pronounced neophobia in mice lacking the PSD protein RICH2
Altered spine morphology and synapse composition in RICH2 mice. a-e Multiple sections from three wild type and knock-out mice each were subjected to Golgi staining (Cerebellum: CER, Cortex: CTX, Hippocampus: HIP, Striatum: STR) (scale bar = 100 μm (CER) and 200 μm CTX, HIP, STR). a–c Analyzing the CA1 region from hippocampal sections reveals no significant changes in spine density between wild type and RICH2−/− mice (a) (unpaired t-test, df = 4, p = 0.68, t =3.18). The spine volume is significantly increased in RICH2−/− mice (b) (unpaired t-test, df = 4, p = 0.0121, t = 3.8407). d Spines were categorized in immature (thin and filopodia like) and mature spines (mushroom, stubby, and multiple spine post-synapses). While no alterations among immature spine types were seen, a shift from mushroom and stubby towards multiple - spine synapses in RICH2−/− mice is visible (1400 (RICH2+/+) and 1459 (RICH2−/−) spines were calculated from at least 10 cells from 3 animals per genotype) (two way factorial ANOVA, genotype F1.44 = 3.013, p = 0.095; morphology F1.44 = 88.471, p < 0.0001, morphology x genotype F1.44 = 7.744, p < 0.0001; RICH2+/+ vs. RICH2−/−: Mushroom: p = 0.0526; Stubby: p = 0.0937; Multiple spine: p < 0.0001; Thin: p = 0.8038; Filopodia like: p = 0.5825) e) Using Scholl-analysis after Golgi staining of hippocampal neurons (10 cells per animal, 3 different animals per group), no significant difference in dendritic arborization between knock-out and wild type mice can be detected (repeated measure ANOVA, genotype F1.114 = 8.398, p = 0.012; distance from soma F1.114 = 1.637, p = 0.022, distance from soma x genotype F1.114 = 1.390, p = 0.089). f Immunohistochemistry of cerebellar sections of wild type (left panel) and Rich2−/− (right panel) mice shows expression of RICH2 among all cell layers (Molecular layer (M), Purkinje layer (P), Granular layer (G)) with increased labeling of Purkinje cells marked by Calbindin immunoreactivity. g Analysis of cerebellar sections using Golgi staining similarly reveals a significantly increased spine volume in RICH2−/− mice (unpaired t-test, df = 4, p = 0.0013, t = 8.0461). h, i Ultra-structural analysis using transmission electron microscopy of brain sections from RICH2+/+ and RICH2−/− cerebellum (h) and hippocampus (i). h) As visible on exemplary images (upper panel, scale bar = 2 μm), the average number of PSDs per optic field of view is significantly increased in RICH2−/− mice (unpaired t-test, Cerebellum: df = 4, p = 0.0044, t = 5.851; Hippocampus: df = 4, p = 0.0106, t = 4.5317). No change in average PSD length and width was observed in cerebellum (analysis was performed using 3 animals per genotype, 12 optic field of view per animal with a total of 581RICH2+/+ and 754 RICH2−/− PSDs analyzed) and in the hippocampal CA1 region (analysis was performed using 3 animals per genotype, 12 optic field of view per animal with a total of 657 RICH2+/+ and 1154 RICH2−/− PSDs analyzed) (unpaired t-test, Cerebellum: length df = 4, p = 0.8622, t = 0.1850; width df = 4, p = 0.6441, t = 0.4989; Hippocampus: length df = 4, p = 0.6298, t = 0.5211; width df = 4, p = 0.3213, t = 1.1309). j Immunohistochemistry, similar to Golgi staining, did not reveal significant changes in synapse density between wild type and RICH2−/− mice in stratum pyramidale (PYR), stratum oriens (OR), and stratum radiatum (RAD) of dendritic spines from hippocampal CA1 pyramidal neurons using unfixed fresh frozen brain sections (14 μm) from wild type and RICH2−/− animals (n = 4 / genotype, 2 analyzed section per animal). Sections were analyzed regarding the number of immunoreactive puncta per optic field (unpaired t-test, df = 4, p = 1.000, t = 2.7764). k Histological immunostainings of CamKIIα/β, GluA1, GluA4, GluN1, Homer1, SHANK2 and SHANK3 in stratum pyramidale (PYR), stratum oriens (OR), and stratum radiatum (RAD) of dendritic spines from hippocampal CA1 pyramidal neurons. Sections were analyzed regarding the average signal intensity per immunoreactive puncta per optic field. A significant increase for GluA4 and GluN1 can be seen in knock-out animals (unpaired t-test, GluA4: df = 18, p = 0.0294, t = 2.3659; GluN1: df = 18, p = 0.0003, t = 4.4109), while the signal intensity and thus protein levels for SHANK3 show a significant decrease in knock-out mice (df = 18, p = 0.0216, t = 2.5151)