Fig 1.
In vivo splicing of exon 9 in the wild type and mutant minigenes.
(a) Schematic representation of the position of the mutations in NF1 exon 9 identified in NF1 patients. (b) Comparison of the transcripts including exon 9 derived from pSPL3-minigenes. Minigenes were transfected into HeLa cells and the total RNA was purified after 24 hours. Transcripts including (FL) and excluding (SK) exon 9 were quantified by semiquantitative RT-PCR and the percentage was calculated with the formula: 100xFL/(FL+SK). All the experiments were performed at least three times (n = 3, ± SD), and the mean value of the percentage for each mutant was normalized to the wild type (WT) minigene (Y-axis). All the mutant minigenes presented a significant reduction in exon 9 inclusion (*** p<0.001), except for the c.1020dupT minigene (p = 0.14).
Fig 2.
Functional analysis of SREs in NF1 exon 9.
(a) Summary of the mutations generated in the E9_pSPL3 minigene. Intronic and exonic sequences are shown in lowercase and uppercase, respectively. The arrows indicate the nucleotide positions substituted and the changes generated. The three putative SREs are shown in grey. Four studies were performed (SRE1, SRE2, SRE3 and the position c.1039) and, for each study, the proportion of transcripts including exon 9 from each mutant minigene was assessed and normalized to the wild type minigene as previously described (Fig 1B). Mutant minigenes that presented a significant reduction in exon 9 inclusion are indicated (*** p<0.001). (b) Study of SRE1. Changes were introduced into the c.908_912 region. (c) Study of SRE2. Changes were introduced into the 943_946 region. (d) Study of SRE3. Changes were introduced into the c.1005_1007 region. (e) Study of position c.1039. Three different changes were tested at c.1039.
Fig 3.
Effect of the strength of the acceptor splice site in exon 9 splicing.
The natural 3’ss of the exon 9 in each minigene was substituted with a more functional one by site-directed mutagenesis (c.889-10_889-9delinsTT). The proportion of transcripts including exon 9 was assessed and normalized to the wild type minigene. The stripped and solid bars represent the exon 9 inclusion from the different minigenes with the natural 3’ss and the 3’ss mutated, respectively. All the differences were statistically significant (*** p<0.001).
Fig 4.
Analysis of SR proteins that regulate the inclusion of exon 9.
Different SR expression constructs cloned in the pGC vector were co-transfected with the pSPL3-minigenes in HeLa cells to test their effect on NF1 exon 9 splicing. (a) Effect of the overexpression of SR proteins on the wild type minigene. The proportion of transcripts including exon 9 was normalized to the assay of HeLa cells co-transfected with the wild-type pSPL3-minigene and the empty pGC vector (control). The overexpression of SRSF2 and SRSF5 enhanced significantly the inclusion of exon 9. (b) Effect of the overexpression of SRSF5 (SRp40) protein on the mutant minigenes. Comparison of the proportion of transcripts containing exon 9 cotransfecting the minigenes with the empty pCG vector (stripped bars) and with the SRSF5 construct (grey bars). The values have been normalized to the wild type assays. The overexpression of SRSF5 enhanced the inclusion of exon 9 in all the mutant minigenes. (c) Effect of the overexpression of SRSF2 (SC35) protein on the mutant minigenes. Comparison of the proportion of transcripts containing exon 9 in experiments co-transfecting the minigenes with the empty pCG vector (stripped bars) or with SRSF2 constructs (grey bars). The values have been normalized to the wild type assays. Just c.945G>A and c.946C>A mutant minigenes failed to augment the inclusion of exon 9 when SRFS2 was overexpresed. (d) Effect of the increase of SRSF2. pSPL3 mutant minigenes were co-transfected with empty pGC (control, stripped bars) and with increasing amounts of SRSF2 constructs (grey bars). The proportions of transcripts containing exon 9 have been normalized to the controls. SRSF2 rising has no effect on the inclusion of exon 9 from c.945G>A and c.946C>A mutant minigenes whereas there is a gradual increase in the inclusion from c.910C>T mutant. (e) Relative quantification of transcripts including exon 9 upon SRSF2 knockdown. The proportions of transcripts from the minigenes are indicated relative to their respective luciferase controls (stripped bars). Transcripts derived from the WT, c.910C>T and c.943C>T minigenes have a significant reduction in exon 9 inclusion (*** p>0.001) while no effect is observed for c.945G>A and c.946C>A mutant minigenes. In the right panel the qPCR results for the SRSF2 expression levels are shown. SRSF2 levels upon SRSF2 siRNA treatment (grey dotted bars) have been normalized to the SRSF2 levels upon luciferase siRNA treatment (white bars) confirming the effective silencing (*** p<0.001).
Fig 5.
Characterization of hnRNPA1 binding to SRE3.
(a) Relative quantification of transcripts including exon 9 upon hnRNPA1 knockdown. The proportion of transcripts is indicated relative to the luciferase controls (stripped bars). The lower panel shows the western blot of the knockdown of hnRNPA1. Tubulin was used as loading control. Silencing of hnRNPA1 increased the inclusion of exon 9 from the c.1006T>A and c.1007G>A mutant minigenes (*** p<0.001). (b) Western blot of the pull-down samples probed with an antibody against hnRNPA1. Four RNA probes of 24 bp spanning the c.1007 position were used to pull-down the proteins: 1007G (wild type), 1007A, 1007T and 1007C (three mutations, respectively). Only the probe containing the c.1007G>A mutations was able to bind to the hnRNPA1 protein. (c) EMSA assay using the recombinant hnRNPA1 protein and the same four probes used for the pull-downs, showing that only this protein binds to the radioactive 1007A probe.