BC RNA Mislocalization in the Fragile X Premutation

Competition of (CGG)₁₀₅ RNA with primate BC200 RNA and rodent BC1 RNA for binding to hnRNP A2. EMSA competition analysis examined binding of human BC200 RNA (A) and rat BC1 RNA (B) to RNA transport factor hnRNP A2. A, Gel was loaded with BC200 RNA (BC200), (CGG)₁₀₅ RNA [(CGG)₁₀₅], and hnRNP A2 (A2) as follows: 1, BC200; 2, BC200 + A2; 3, (CGG)₁₀₅; 4, (CGG)₁₀₅ + A2; 5, BC200 + (CGG)₁₀₅; 6–8, BC200 + A2 + (CGG)₁₀₅ at 2, 5, or 10 min of incubation. B, Gel was loaded with BC1 RNA (BC1), (CGG)₁₀₅, and A2 as follows: 1, BC1; 2, BC1 + A2; 3, (CGG)₁₀₅; 4–6, BC1 + A2 + (CGG)₁₀₅ at 10, 20, or 30 min of incubation time. Components were used at 100 nm (Muslimov et al., 2011) except for (CGG)₁₀₅ RNA, which was used at a 1:3 molar ratio to the respective BC RNA.

Impaired dendritic localization of BC1 RNA in (CGG)₁₈₀ animals. CA1 strata oriens (O), pyramidale (P), radiatum (R) are indicated. A, B, Coronal sections through hippocampal regions (film autoradiograms) reveal that BC1 RNA (white signal) is concentrated in CA1 strata oriens and radiatum (i.e., dendritic layers) of WT animals but, in contrast, is concentrated in stratum pyramidale (i.e., cell body layer) of CGG animals. C, D, Emulsion autoradiography of the CA1 hippocampal region confirms predominantly dendritic BC1 labeling (white signal) in WT CA1 but predominantly somatic labeling in CGG CA1. Scale bar for C and D: 100 μm. E, F, Quantitative analysis of emulsion autoradiographs: one-way ANOVA, p < 0.001. Dunnett’s post hoc analysis, comparison of signal intensities (given in relative units) in stratum P (center), in stratum O at distances of 50 and 100 μm from edge of stratum P, and in stratum R at distances of 50, 100, 150, and 200 μm from edge of stratum P, between WT and CGG animals: p < 0.001 for all sample points. n = 4 for WT and CGG animals. Error bars indicate SEM.

Somato-dendritic distribution of MAP2 mRNA in WT and CGG brains. A, B, MAP2 mRNA distribution in hippocampal CA1 of WT and CGG animals. CA1 strata oriens (O), pyramidale (P), radiatum (R) are indicated. Scale bars, 100 μm. C, D, Quantitative analysis: one-way ANOVA, p = 0.94065. Comparison of signal intensities (given in relative units) in stratum P (center), in stratum O at distances of 50 and 100 μm from edge of stratum P, and in stratum R at distances of 50, 100, 150, and 200 μm from edge of stratum P, between WT and CGG animals: p > 0.05 for all sample points. n = 4 for WT and GCC animals.

Prolonged epileptiform discharges in CGG CA3 pyramidal cells. The GABA_A receptor antagonist bicuculline induced group I mGluR-mediated prolonged epileptiform discharges in CA3 hippocampal pyramidal cells. A, Intracellular recordings of the spontaneous activity of a CA3 pyramidal cell in a CGG hippocampal slice preparation (upper panels) and of a CA3 pyramidal cell in a WT hippocampal slice preparation (lower panels) after perfusion with bicuculline (50 μm). Within 30 min after addition of bicuculline, short synchronized discharges were elicited in both preparations (left). Continuous application of bicuculline induced prolonged synchronized discharges (4–7 s) in CGG animal preparations but not in WT animal preparations (middle). Addition of group I mGluRs antagonists LY367385 and MPEP (80 μm) reversed prolongation of synchronized discharges in CGG animals (right). B, Frequency histograms of all synchronized bursts recorded during a 6-min period of stable rhythmic activity at three time points: Bic 30 min; Bic 90 min; LY367385 + MPEP 60 min. C, Summary bar graph of average burst durations in CGG and WT preparations. The average burst duration in CGG preparations at Bic 90 min was significantly higher than that observed in WT preparations (5.106 ± 0.148 s vs. 0.504 ± 0.01 s; one-way ANOVA, post hoc Tukey HSD test, p < 0.001).

Audiogenic seizures in CGG animals. Significantly increased propensity for audiogenic seizures was observed with CGG mice (n = 42), in comparison with WT mice (n = 30). Seizures were not observed in CGG mice injected with anisomycin (75 mg/kg i.p.; n = 12) or MPEP (40 mg/kg i.p.; n = 10). Fisher’s exact test, p < 0.001. Error bars represent 95% confidence intervals.

Self-grooming of CGG animals. CGG mice spent significantly more time self-grooming than WT mice (Mann–Whitney test, p = 0.004). n = 13 for each group.

Cognitive flexibility is impaired in CGG animals. Numbers of ETC (A, C, E, and G) and TTC (B, D, F, and H) were recorded. A, B, In Phase 1, CGG animals were significantly impaired, in comparison to WT mice, in sessions SD learning, CD Learning, and Conflict Learning. C, D, CGG animals and WT animals performed similarly in Phase 2 sessions CD Learning 2 and Conflict Learning 2. E, F, In Phase 3, CGG and WT animals performed comparably in session CD Learning 3 but CGG animals displayed continued cognitive impairment in session Conflict Learning 3. G, H, In Phase 4, no significant differences were observed in the performances of CGG and WT animals in session CD Learning 4 and Conflict Learning 4. EDS, extradimensional shift; IDS, intradimensional shift. n = 9 for each animal group (CGG and WT).

BC1 RNA DTE and CGG-repeat stem-loops: noncanonical motif structures. Noncanonical purine∙purine pairs are symbolized by ∙, standard WC pairs by = (GC) or – (AU), wobble WC pairs by ·. In the BC1 RNA DTE, the noncanonical GA core motif resides in an A-form helix that is part of the 5′ BC1 apical stem-loop domain (Muslimov et al., 2006; 2011). The GA core (red) is clamped by canonical base pairs which are mostly G=C standard WC (blue). The structure of the 5′ BC1 domain was established experimentally (Rozhdestvensky et al., 2001). In CGG-repeat stem-loops, noncanonical G∙G pairs (red) are flanked by G=C standard WC pairs (blue; Napierala et al., 2005; Zumwalt et al., 2007; Kiliszek et al., 2011). Noncanonical R∙R pairs (e.g. A∙G, G∙G) are rather strong, comparable in stability to A-U WC pairs (Mládek et al., 2009; Sobczak et al., 2010).