Figure 1. Specificity of the chemical and enzymatic probes used.

Figure 2. Analysis by UV‐crosslinking assays of the competition between SRSF1 and hnRNP A1 proteins for binding to a HIV fragment RNA fragment involved in splicing regulation. RNP complexes were formed on a HIV RNA fragment involved in splicing regulation using either SRSF1 or hnRNP A1 recombinant proteins alone (0: 1 or 1: 0, respectively) or both proteins in equimolar amounts (1: 1). The RNPs were UV‐irradiated at 254 nm. Free and crosslinked SRSF1 and hnRNP A1 proteins were immunoselected on beads using specific antibodies directed against each protein [hnRNP A1 (4B10) and SRSF1 (1D7)], and fractionated on 10% SDS‐ PAGE. The total amounts of immunoselected proteins were estimated by Western blot analyses, and the yields of protein crosslinking compared by autoradiography. When present alone at a 3 µM concentration, each protein binds to the RNA. When the two proteins are present together at a 3 µM concentration, hnRNP A1 binds to the RNA but not to SRSF1, thus demonstrating a competition of the two proteins to bind to the same splicing regulatory site. When hnRNP A1 is bound, the splicing is inhibited; when SRSF1 is bound, the splicing is activated.

Figure 3. Footprinting analysis of (CUG) ₁₆/MBNL1 (a and b) and (CUG) ₁₆/CUG‐BP1 (c and d) complexes. About 10 fmol of 5′ end‐labeled (CUG) ₁₆ transcript were incubated in the absence (–) or the presence of 20 or 60 pmol of MBNL1 (panel a) or CUG‐BP1 (panel c) recombinant protein. The complexes formed were digested with the RNases indicated at the top of the autoradiogram. The digestion products were fractionated by electrophoresis on a 7% sequencing gel. The lane OH ⁻, corresponds to an alkaline hydrolysis performed under denaturing conditions; this was used for localization of the cleavage sites. G positions in the (CUG) ₁₆ RNA are indicated on the left side of the autoradiogram. In panel (b), the protections observed in the (CUG) ₁₆/MBNL1 complex are represented by blue circles on the secondary structure deduced for (CUG) ₁₆ RNA. Arrows indicate the part of the RNA which was analyzed. Most of the G residues in the helix are protected, as well as all residues in the single‐stranded terminal loop, showing the multimerization of MBNL1 all along the stem–loop structure. In contrast, no protection is detected in the presence of CUG‐BP, showing the absence of binding of this protein on CUG repeats. This explains why, in the nucleus from patients suffering from myotonic dystrophy type 1, the amplified CUG repeats present in the DMPK pre‐mRNA capture the splicing regulatory factor MBNL1, but not the splicing regulatory factor CUG‐BP. This leads to a strong deregulation of alternative splicing.

Figure 4. Supershift assays performed on an RNP complex formed by the incubation of HIV‐1 RNA fragments involved in splicing regulation in a HeLa cell nuclear extract. The HIV‐1 RNA fragment was produced by in vitro transcription and was 5′ end‐labeled. An RNP complex was formed by its incubation in HeLa cell nuclear extract. To investigate the presence of hnRNP A1 in this complex, a supershift assay was performed using anti‐hnRNP A1 antibodies. The free RNA, the RNA incubated with the antibody, and the RNP complex formed in nuclear extract without or with further addition of the antibody, were fractionated by electrophoresis on a nondenaturing 6% polyacrylamide (38: 2) gel in 1 mM EDTA, 45 mM Tris borate (pH 8.3) buffer. The positions of the free RNA, the RNP, and the supershifted RNP are indicated on the left‐hand side of the autoradiogram. These data demonstrate the presence of the splicing inhibitor hnRNP A1 in the splicing regulatory complex.