Fig 2: The transcriptional stop cassette was inserted downstream of Ube3a via homologous recombination.(A) The murine genomic structure of the PWS/AS region is shown (not to scale), with paternally expressed genes in blue, maternally expressed genes in red, and biallelically expressed genes in grey. Genomic regions deleted in the knock-out alleles of del0.9, del4.8, delS-U, delU-G, and Ube3aKO are marked by the green bars. Insertion site of the poly(A) cassette in Ube3aATS-stop allele is shown by the “STOP” sign. (B) The mouse wild-type locus is shown with enlargement of the region between Snord115 and Ube3a. The targeting vector contains the poly(A) cassette and the neomycin selection marker. The slanted boxes are left and right probes used in the Southern blot analysis. LA, left arm; RA, right arm; B, BglII; red arrows, genotyping primers; green arrows, q-PCR primers of Ube3a-ATS 5' and 3'; triangles, loxP sites. (C) Recombinant ES cell clones were identified by Southern blot with BglII digesion. The WT allele gives a band of 14.6 kb and the knock-in allele giving a band of 8.7 kb, hybridizing to either LA or RA probe. (D) Chimeric mice were confirmed by multiplex PCR, with WT allele of 742 bp and mutant allele of 358 bp.

Fig 3: AS TR2-UBE3A mice had significant improvements in the Morris water maze.(A) Escape latency to reach the platform during 5 days of training in Morris water maze. The only significant differences seen were an increase in latency for the AS TR2-GFP mice on day 3 and a decrease in latency for wildtype mice on day 4 compared to the other two groups (2-way ANOVA Bonferroni: Interaction [F(8,100) = 1.01, P>0.05]; Treatment [F(2,100) = 5.30, P<0.05]; Time [F(4,100) = 53.49, P<0.0001]; Matching [F(25,100) = 5.37, P<0.0001]). The target platform is indicated by the black circles. (B) Quantification of the number of platform crossings in the target (T), opposite (O), right (R), and left (L) quadrants during the probe trial of the Morris water maze 24 hours after training indicate no significant differences among any of the groups in comparing target platform crossings to opposite platform crossings (ANOVA Tukey WT: [F(3,35) = 9.546, P<0.0005]; AS TR2-GFP: [F(3,35) = 3.186, P<0.01]; AS TR2-UBE3A: [F(3,39) = 2.814, P<0.05]). All three groups learned the platform location based on a spatial bias as indicated by the time spent in the target quadrants (Fig. 5D 24 h ANOVA Tukey [F(2,26) = 1.027, P>0.05]) (C) A probe test 72 hours after training indicate that the WT and AS TR2-UBE3A groups had significantly more target platform crossings compared to the number of opposite platform crossings (ANOVA Tukey WT: [F(3,35) = 6.086, P<0.005]; AS TR2-GFP: [F(3,35) = 0.5650, P>0.05]; AS TR2-UBE3A: [F(3,39) = 2.679, P<0.05]), but this improvement was not spatially biased as seen by the time spent in the target quadrant (Fig. 5D 72 h ANOVA Tukey [F(2,25) = 5.067, P<0.02]). Results shown represent the mean with standard error.

Fig 4: Arc and Ube3A do not interact in the hippocampus in vivo. (A) Arc was immunoprecipitated from whole hippocampal lysate from P30 WT mice and immunoblotted for Arc and Ube3A. Arc and Ube3A do not coimmunoprecipitate. (B) Hippocampal lysates from P30 WT mice underwent subcellular fractionation and Arc was immunoprecipitated from the nuclear and synaptoneurosome fractions. Arc and Ube3A do not coimmunoprecipitate. I, input; IB, immunoblot; nuc, nucleus; synapto, synaptoneurosome; U, unbound fraction.

Fig 5: Transcriptional collision hypothesis of Ube3a imprinting.Paternal Ube3a was found to be transcribed at a similar level as maternal Ube3a at the 5' portion of the gene. The levels of both paternal Ube3a sense and antisense transcripts were found to be decreased around intron 4 of Ube3a. Based on these observations, a hypothesis of transcriptional collision is proposed in which the two opposing polymerases of Ube3a and Ube3a-ATS on the paternal chromosome collide into each other around intron 4 of Ube3a. The collision causes the polymerases to stall or fall off from the templates, thereby completely or partially terminating the transcription of Ube3a and Ube3a-ATS. The incomplete sense transcript of Ube3a is then degraded and unable to be processed into full-length mature mRNA.