Figure 1.
Generation of a null Khdrbs3 allele.
(A) Northern analysis of different adult mouse tissues to detect expression of Khdrbs3 (upper panel) and small subunit rRNA (lower panel). (B) The genomic structure of the Khdrbs3 alleles from wild type, floxed, and null mice mice were monitored using Southern blotting and the probe indicated in parts C–D. The Southern blot demonstrates that the cross with a PGK-Cre mouse successfully removed exon 2 from the genomic DNA. (C) Genomic structure of the Khdrbs3LoxP conditional allele in which exon 2 of the Khdrbs3 gene is flanked by LoxP sites. (D) Genomic structure of the null (Khdrbs3−) allele from which exon 2 has been deleted by Cre-mediated recombination. (E) Multiplex RT-PCR analysis of Khdrbs3 and Hprt mRNA levels in different mouse tissues. The size markers are shown in nucleotides. (F) Western blot analysis of Sam68 and T-STAR protein levels in the testes of wild type and Khdrbs3 null mice using an antibody that recognizes T-STAR and Sam68. The position of the size markers are shown in KDa.
Figure 2.
T-STAR is not required for male fertility in the mouse.
(A) Histology of testis from wild type and Khdrbs3−/− animals. Paraffin-embedded sections were probed for T-STAR using a T-STAR specific antibody or Sam68 using a Sam68 specific antibody and counterstained using haematoxylin. The scale bar is equivalent to 20 µm. (B) Litter sizes obtained for HET (n = 40 litters) and Khdrbs3−/− (n = 23 litters) mice. Slightly smaller than average litter sizes were observed for the KO mice. (C) Testis-body weight ratios were not significantly different between WT (Khdrbs3+/+), HET and KO mice indicating no significant defect in adult testis development in the absence of T-STAR protein on a C57BL6 background or a mixed C57Bl6-129 background. (D) Sperm counts of wild type (n = 4); HET (n = 10) and KO (n = 9) mice indicate no reduction in sperm count associated with the knockout allele. (E) Mendelian ratios of litters born from breeding HET mice correspond to the expected 1∶2∶1 ratio, and so indicate no lethality associated with the KO allele (n = 40 litters).
Figure 3.
T-STAR protein is a dose-dependent splicing regulator of Nrxn1-3 AS4 in the mouse brain.
(A) Agarose gels showing levels of AS4 splicing inclusion from each of the Nrxn1, Nrxn2 and Nrxn3 genes using three different mice from each genotype. (B) Percentage Splicing Exclusion levels of Nrxn1-3 AS4 and Stxbp5l exon 23 measured measured by RT-PCR and capillary gel electrophoresis in wild type (WT), Khdrbs3+/− (HET) and Khdrbs3−/− mice (KO) (n = 3 mice for each genotype). The p values were calculated from two tailed unpaired t tests, and error bars represent standard errors of the mean.
Figure 4.
T-STAR protein regulates region-specific splicing of Neurexin1-3 AS4 in the mouse brain.
(A) Neurexin splicing regulation in different regions of the mouse brain (B) Schematic of the different mouse brain regions used for analysis. (C–E) Percentage splicing exclusion in different regions of the mouse brain (n = 3 mice from each genotype) measured in RNA samples from wild type (column +) and knockout (column −) mice for AS4 of (C) Nrxn1, (D) Nrxn2 and (E) Nrxn3. The error bars correspond to the standard error of the mean. Statistical significances were calculated using a two tailed unpaired t test. No splicing exclusion was observed for Nrxn3 in the absence of T-STAR protein in any brain region.
Figure 5.
T-STAR protein is expressed in the embryonic brain.
(A) Section of embryonic brain (13.5 day embryo) including most of the forebrain region, stained for T-STAR (brown) and counterstained with haematoxylin (blue). (B) Nrxn1-3 exon AS4 splicing patterns in wild type and T-STAR knockout 13.5d embryonic brain.
Figure 6.
T-STAR protein concentration correlates with Nrxn1-3 AS4 alternative splicing patterns.
(A) Expression levels of T-STAR and Sam68 protein in different regions of the mouse brain were measured using Western blotting. The same filters were first probed with antisera specific for T-STAR, and then stripped and reprobed with an antisera specific for Sam68. (B–D) Levels of Nrxn1-3 AS4 Percentage Splicing Exclusion in each brain region plotted against the ratio of T-STAR: Sam68 protein quantified from the Western blot shown in (A). The dashed line is the 95% confidence limit of the best fit line.
Figure 7.
Nrxn exon AS4 alternative splicing control is dependent on the physiological expression of T-STAR protein even though Sam68 is co-expressed.
(A) Immunolocalisation of T-STAR and Sam68 proteins in the mouse hippocampus from wild type or knockout mouse brains (Abbreviations: DG - Dentate Gyrus; and AH -Ammon's Horn). The scale bar is equivalent to 20 µm). (B) Immunolocalisation in the mouse testis. Paraffin embedded adult mouse testis sections were stained with affinity purified antibodies raised against T-STAR or Sam68 (brown staining), and counterstained with haematoxylin (blue). Abbreviations: Spg –spermatogonia; Spc –spermatocyte; Rtd –round spermatid; Spd –elongating spermatid; SC –Sertoli cell. The size bar corresponds to 20 µM. (C) Levels of Nrxn1 and Nrxn3 AS4 alternative splice isoforms in the testes of different mouse genotypes (n = 3 mice of each genotype) measured by RT-PCR and agarose gel electrophoresis. (D) Quantification of Percentage Splicing Exclusion in the testes of different mouse genotypes using capillary gel electrophoresis (n = 3 mice of each genotype: wild type mice Khdrbs3+/+ (abbreviated WT) Khdrbs3+/− mice (abbreviated HET) and Khdrbs3−/− mice (abbreviated KO). The p values were calculated using unpaired t tests, to determine the significance of the difference between percentage splicing exclusion levels in the wild type versus either the heterozygous Khdrbs+/− mice (HET); or wild type versus the homozygous Khdrbs3−/− (KO) mice. The standard error of the mean is shown as an error bar.
Figure 8.
Nrxn2 is a specific splicing target for T-STAR but not Sam68.
(A) Comparative modular organisations of the Sam68 and T-STAR proteins. The position of the serine residue 20 (S20) in Sam68 which is phosphorylated to mediate splicing regulation of Nrxn1 AS4 is indicated. (B) Representative Western blot analysis of HEK293 cells co-transfected with expression constructs for GFP-fusion proteins and minigenes. The Western blot was probed for GFP and actin, showing equal expression levels of each fusion protein. (C) Nrxn2 AS4 is regulated by T-STAR but not Sam68. Representative capillary gel electrophoretogram (top) and bar chart (bottom) showing splicing data from single and three independent transfections respectively. (D) Nrxn3 splicing is regulated by T-STAR and Sam68. Representative capillary gel electrophoretogram (top) and bar chart (bottom) showing splicing data from single and three independent transfections respectively. In each case, statistical significance was compared between HEK293 cells expressing GFP (lane1; shown as green bar) and HEK293 cells expressing the GFP fusion proteins (lanes 2–8 shown as black bars) using a non-paired t test and the error bar represents the standard error of the mean.
Figure 9.
T-STAR mediates splicing repression of Nrxn2 AS4 via a downstream response element.
(A) Nucleotide sequence immediately downstream of the Nrxn2 alternative exon (exon sequence shaded grey, intron sequence unshaded) of wild type gene and after mutagenesis to remove U((U/A)AA motifs. (B) Capillary gel electrophoretogram to show splicing response of mutated Nrxn2 minigene after co-transfection with the indicated expression constructs. (C) Percentage Splicing Exclusion levels of the mutated Nrxn2 minigene from 3 independent transfection experiments. The p values were calculated using unpaired t tests, and estimate statistical significance between splicing exclusion of the mutated Nrxn2 minigene on co-expression of GFP(lane 1: shown as green bar) and each of the individual fusion proteins (lanes 2,4–7: shown as black bars). (D) Coomassie stained gel showing purified Sam68-GST and T-STAR-GST fusion proteins used for gelshift experiments. (E) EMSA experiment using RNA probe containing the Nrxn2 response element. (F) EMSA experiment using a control RNA probe.
Figure 10.
Distinct arrangements of UWAA motifs are shared between Neurexin gene paralogs in bony vertebrates and predict an ancient mechanism of splicing control.
Annotation of UWAA motifs in the 200 nucleotide regions downstream of AS4 in human (abbreviated Hs); tortoise (abbreviated Ps); Ceolocanth (abbreviated Lc); and zebrafish (abbreviated Dr). The Nrxn genes have duplicated in the zebrafish, and these two copies are shown distinguished as the Nrxn1-3a and Nrxn1-3b copies. Nucleotides are shown as red (A), green (U), blue (C) and yellow (G). Concentrations of UWAA motifs are boxed as indicated in the key.
Figure 11.
Human T-STAR protein represses splicing inclusion of the Zebrafish Nrxn3 AS4 exon.
(A) Capillary gel electrophoretogram showing splicing of a minigene encoded zebrafish Nrxn3 in response to co-expressed proteins introduced by co-transfection. (B) Quantification of biological replicates from three independent co-transfection experiments.
Figure 12.
T-STAR null mice have long and short term spatial memory equivalent to wild-type mice.
(A) Wild type and knockout mice were trained each day(four trials) for four days in a Barnes maze, and the time taken to find the escape hole (primary latency) was monitored. (B) Time taken to find the escape hole on day 5 (short term memory) and on day 12 (long-term memory)with no training between day 5 and day 12 (n = 8 wild type and 11 knockouts and error bars indicate the standard error of the mean).
Figure 13.
Concentration-dependent splicing model for regional regulation of Nrxn1-3 AS4 in the mouse brain.
T-STAR protein directly regulates Nrxn1-3 AS4 splicing. In the cerebellum T-STAR concentrations are low and most of the Nrxn1-3 mRNA isoforms include AS4 as a result. In the forebrain-derived regions T-STAR concentrations are high, and there are also increased levels of AS4 splicing exclusion resulting from this. Sam68 protein levels are similar across the brain regions.