The non-coding RNA BC1 regulates experience-dependent structural plasticity and learning

Abstract

The brain cytoplasmic (BC1) RNA is a non-coding RNA (ncRNA) involved in neuronal translational control. Absence of BC1 is associated with altered glutamatergic transmission and maladaptive behavior. Here, we show that pyramidal neurons in the barrel cortex of BC1 knock out (KO) mice display larger excitatory postsynaptic currents and increased spontaneous activity in vivo. Furthermore, BC1 KO mice have enlarged spine heads and postsynaptic densities and increased synaptic levels of glutamate receptors and PSD-95. Of note, BC1 KO mice show aberrant structural plasticity in response to whisker deprivation, impaired texture novel object recognition and altered social behavior. Thus, our study highlights a role for BC1 RNA in experience-dependent plasticity and learning in the mammalian adult neocortex, and provides insight into the function of brain ncRNAs regulating synaptic transmission, plasticity and behavior, with potential relevance in the context of intellectual disabilities and psychiatric disorders.Brain cytoplasmic (BC1) RNA is a non-coding RNA that has been implicated in translational regulation, seizure, and anxiety. Here, the authors show that in the cortex, BC1 RNA is required for sensory deprivation-induced structural plasticity of dendritic spines, as well as for correct sensory learning and social behaviors.

Fig. 1

BC1 KO neurons have increased spine density and decreased dendritic complexity. a Representative images of WT and BC1 KO cultured primary neurons transfected with EGFP. Scale bar, 100 µm. b Sholl analysis and quantification of the number of intersections as a function of distance from the soma (***P < 0.001, two-way ANOVA, mean ± s.e.m., n = 37 WT and 36 KO neurons). c Dendritic length in WT and BC1 KO neurons (*P < 0.05, two-tailed t-test, mean ± s.e.m., n = 37 WT and 36 KO neurons). d left Representative images of dendritic segments of neurons transfected with EGFP. Scale bar, 2 µm. right Quantification of spine density in EGFP-transfected cortical neurons (***P < 0.001, two-tailed t-test, mean ± s.e.m., n = 76 and 68 dendritic segments from WT and KO neurons, respectively). e Scheme of the cortical layers and dendrites analyzed in the somatosensory cortex relative to f–h. f–h left In each group, representative images of dendritic segments of neurons from layers 2/3, 4 or 5 used to analyze spine density. right Median values ± median absolute deviation (m.a.d.) of spine density distributions (***P < 0.001, Mann–Whitney U-test; n = 136/187, 201/218 and 137/180 dendritic segments from layers 2/3, 4 and 5, respectively, in WT/KO mice). Scale bar, 5 µm

Fig. 2

BC1 KO pyramidal neurons have enlarged spine heads and postsynaptic densities. a Biocytin-filled dendritic segments from layer 2/3 barrel cortex neurons. Scale bar, 2 µm. b Frequency distribution of spine head size in dendritic segments (χ ² = 32.99, df = 1, ***P < 0.001, n = 1303 and 2747 spines for WT and KO mice, respectively). c Representative electron micrographs of spines from the barrel cortex of WT and BC1 KO mice. Scale bar, 250 nm. d Frequency distribution of spines as a function of PSD length (n = 369 and 615 spines for WT and BC1 KO mice, respectively). e Representative electron micrographs after phosphotungstic acid staining depicting the postsynaptic density (PSD) and active zone (AZ) of spines from the barrel cortex of WT and BC1 KO mice. Scale bar, 100 nm. f, g Mean ( ± s.e.m.) PSD length and thickness in BC1 KO compared to WT mice (**P < 0.01, ***P < 0.001, Mann–Whitney U-test, n = 99 and 83 spines for WT and KO, respectively). h Three-dimensional reconstruction of synapses from the barrel cortex of WT and BC1 KO mice after SBF-SEM imaging. Dendritic spines (blue), PSD (yellow), presynaptic terminals (green), AZ (orange) and synaptic vesicles (magenta) are illustrated. Scale bar, 500 nm

Fig. 3

BC1 KO pyramidal neurons display increased spontaneous synaptic activity. a Sample traces of spontaneous EPSCs (sEPSC) recorded from layer 2/3 pyramidal neurons of WT and BC1 KO barrel cortex. Scale bars, 20 pA, 100 ms. b–d Quantification of sEPSC amplitude, frequency and kinetics (**P < 0.01, two-tailed t-test, mean ± s.e.m, n = 13 and 17 neurons from WT and KO mice, respectively). Outliers from b and d (closed circles) were excluded from the statistical analysis. n.s., not significant. e Scheme illustrating the stimulation of principal (PW) and surrounding whiskers (SW). f Spontaneous activity in layer 2/3 barrel cortex neurons (*P < 0.05, unpaired t-test with Welch correction, mean ± s.e.m., n = 4 and 10 neurons from WT and KO mice, respectively). g, h Stimulation of PW and SW and response rates of layer 2/3 pyramidal neurons (P = 0.4141 and 0.4201, unpaired t-test with Welch correction, mean ± s.e.m., n = 4 and 10 neurons from 3 WT and 5 KO mice, respectively). i left Photomicrographs of cytochrome oxidase (CO)-stained barrels in somatosensory cortex of WT and BC1 KO mice. Scale bar, 500 µm. right Scatter plot of the mean optical density of CO staining in the barrels. Dots represent individual barrel values (***P = 0.001, Mann–Whitney U-test, mean ± s.e.m., n = 18 barrels)

Fig. 4

BC1 KO mice show increased synaptic levels of GluRs and PSD-95. a left Representative western blotting of different GluR subunits and synaptic proteins in PSD-enriched fractions from WT and BC1 KO mice. right Quantification of protein levels; values are expressed as ratio (fold of WT) of protein levels over actin (*P < 0.05, **P < 0.01, two-tailed t-test, mean ± s.e.m., n = 6 WT and 9 KO mice). b top Representative immunoblots of phospho-αCaMKII (Thr286) and total αCaMKII in PSD-enriched fractions from WT and BC1 KO mice. bottom Quantification of protein levels; values are expressed as ratio (fold of WT) of protein levels over actin (*P < 0.05, two-tailed t-test, mean ± s.e.m., n = 6 WT and 9 KO mice). c–e top Representative western blotting of the indicated GluRs in WT and BC1 KO synaptoneurosomes following 1 h treatment with vehicle (Cnt), 4EGI (50 µM), or cycloheximide (CHX, 25 µM). bottom Quantification of GluR over actin levels (*P < 0.05 vs. WT Cnt, ^# P < 0.05 vs. KO Cnt, two-way ANOVA, mean ± s.e.m., n = 6–13 mice). c WT Cnt (n = 9), WT 4EGI (n = 7), WT CHX (n = 7), KO Cnt (n = 8), KO 4EGI (n = 8), KO CHX (n = 8). d WT Cnt (n = 13), WT 4EGI (n = 6), WT CHX (n = 6), KO Cnt (n = 11), KO 4EGI (n = 11), KO CHX (n = 11). e WT Cnt (n = 10), WT 4EGI (n = 7), WT CHX (n = 7), KO Cnt (n = 7), KO 4EGI (n = 7), KO CHX (n = 6)

Fig. 5

Translation of PSD-95 mRNA is dysregulated in BC1 KO synaptoneurosomes. a left Representative western blotting showing incorporated puromycin in WT and BC1 KO cortical neurons. right Quantification of puromycin immunodetection normalized to Ponceau red staining (***P < 0.001 vs. WT, ^### P < 0.001 vs. KO, one-way ANOVA, mean ± s.e.m., n = 9 WT, 16 KO (Cnt), 6 KO (4EGI) and 8 KO (CHX) mice). b Synaptoneurosomes from WT and BC1 KO mice were incubated with vehicle (Cnt) or 100 µM DHPG for 10 min in presence of L-AHA. Representative western blotting for all de novo-synthesized (biotin-positive) proteins detected using IRDye 800CW Streptavidin. Blots are different parts of the same gel (see Supplementary Fig. 6 for full scans). c left Representative western blotting showing the levels of ubiquitinated proteins in WT and KO synaptoneurosomes. right Quantification of ubiquitinated proteins normalized to Ponceau red staining (P = 0.4202; two-tailed t-test, mean ± s.e.m., n = 6 WT and 7 KO mice). n.s., not significant. d left Western blotting for de novo-synthesized (biotinylated) and total PSD-95, after immunoprecipitation (IP) with PSD-95 antibody. right Quantification of immunoblots; data are presented as the ratio (fold of WT-control) of biotinylated PSD-95 over total PSD-95 (*P < 0.05, two-way ANOVA, mean ± s.e.m., n = 6 WT and 5 KO mice). e Levels of BC1 RNA and of PSD-95 and αCaMKII mRNAs in cortical extracts from WT (white bars) and BC1 KO (black bars) mice. Values were normalized for HPRT and Gusb mRNA levels and expressed as fold of WT (mean ± s.e.m., n = 5 WT and 3 KO mice)

Fig. 6

BC1 KO mice show abnormal spine plasticity in response to whisker deprivation. a Schematic illustrating the cortical layers and dendrites of the somatosensory cortex analyzed 1 week after whisker deprivation. The brain slice photograph was taken from MBL brain atlas of the C57BL/6j mouse (http://www.mbl.org/atlas170/atlas170_frame.html). b Changes in spine density after whisker deprivation in WT and BC1 KO mice were detected by Golgi staining and are expressed as percentage relative to their respective control groups (*P < 0.05 as compared to control, Mann–Whitney U-test, mean ± s.e.m., layer 2/3: n = 15/14 and 14/13 neurons for I/C WT and KO mice, respectively; layer 4: n = 12 and 13 neurons for WT and BC1 KO mice, respectively). c Representative immunoblots (left) and quantitative analysis (right) of the levels of GluR subunits in the ipsilateral [Ipsi] and contralateral [Contra] barrel cortex of WT and BC1 KO mice 1 week after whisker deprivation. Values expressed as ratio (fold of ipsilateral) of GluR over Vinculin protein levels (*P < 0.05, two-tailed t-test, mean ± s.e.m., n = 9 WT and 7 KO mice)

Fig. 7

BC1 KO mice have impaired texture novel object recognition (tNORT) and social behavior. a, b Preference index (novel vs. familiar object) in the visual novel object recognition test (vNORT) and tNORT (*P < 0.05, **P < 0.01, ***P < 0.001 vs. chance level, one-sample t-test, mean ± s.e.m., n = 11 WT and 10 KO mice). c, d Preference index for sociability (stranger 1 vs. empty cage) and for social novelty (stranger 2 vs. stranger 1) in the three-chamber test (*P < 0.05, **P < 0.01, ***P < 0.001 vs. chance level, one-sample t-test, mean ± s.e.m., n = 17 WT and 11 KO mice). e, f Time to reach the goal during the training sessions and percentage of matches won in the automated tube test (n = 8 mice). g, h Time spent self-grooming and number of marbles buried by WT and BC1 KO mice (P = 0.1516 and 0.9336, respectively, two-tailed t-test, mean ± s.e.m., n = 17 WT and 11 KO mice). i, j Nest-building test score and unused cotton material (P = 0.5313 and 0.7289, respectively, two-tailed t-test, mean ± s.e.m., n = 9 WT and 11 KO mice). n.s., not significant

Fig. 8

BC1 RNA-mediated regulation of structural plasticity at cortical synapses. In WT mice, PSD-95 mRNA translation is repressed through multiple mechanisms including regulation by FMRP and miR125a , association of BC1 RNA with the FMRP-CYFIP1 complex^, , and interaction between BC1 RNA and eIF4A/B^, . These different ribonucleoprotein complexes mediate activity-regulated translational control of PSD-95 mRNA. Absence of BC1 RNA causes exaggerated local synthesis of PSD-95 and enhanced synaptic delivery of ionotropic glutamate receptors to the postsynaptic density (PSD). These biochemical alterations along with the oversized PSD and spine heads might contribute to the increase in synaptic transmission and to the behavioral abnormalities observed in BC1 KO mice. The figure was produced using Servier Medical Art (http://www.servier.com)