In vitro splicing of yeast actin branchpoint mutants promoted by an RS domain. (A) A chimeric protein in which the MS2 RNA-binding domain was fused to the RS domain of ASF/SF2 was analyzed for its ability to support splicing of the wild-type actin-MS2(E2) pre-mRNA substrate or mutant derivatives (shown on the left; the branchpoint adenine is underlined). In vitro splicing reactions were performed in yeast WCE to which no protein (−) or recombinant, purified MS2 or MS2–(ASF)RS protein was added, as indicated. Identities of the spliced products are shown on the right. (B) MS2–(U2AF⁶⁵)RS and MS2–(U2AF³⁵)RS were analyzed for their ability to rescue splicing of the actin-MS2(E2)/ 1U-A and actin-MS2(E2)/2A-U mutant substrates. (C) UV cross-linking analysis of the RS domain–pre-mRNA interaction with the wild-type actin-MS2(E2) (top), actin-MS2(E2)/1U-A (middle), and actin-MS2(E2)/5A-U (bottom) substrates labeled with a single ³²P-G at the 5′ splice site (5′ss), branchpoint (BP), Py-tract (Py), 3′ splice site (3′ss), or MS2-binding site (MS2), indicated in the schematic diagram (bottom). Molecular weight markers are shown on the left in kilodaltons, and the positions of the 26-kDa ³²P-tagged MS2–TF–(ASF)RS and 8-kDa ³²P-tagged (ASF)RS TEV-cleaved product are indicated on the right.

In vivo splicing of yeast actin branchpoint and 5′ splice site mutants promoted by an RS domain. Splicing of the pre-mRNAs was measured by an RNase-protection assay in yeast strains expressing a wild-type or mutant actin-MS2(E2) (left) or actin-MS2(I) pre-mRNa (right) derivative. Each strain also expressed an MS2–(ASF)RS domain fusion protein, the control MS2 protein, or no MS2 derivative. RNase protection assays used a probe that spanned the exon–intron junction (shown below), which allowed for differentiation of the spliced and unspliced products. Identities of the unspliced (pre-mRNA) and spliced (mRNA) products are indicated on the right. The amount of spliced product was quantitated relative to the amount of product obtained with a wild-type pre-mRNA substrate in the absence of an MS2 fusion protein (percent wild type: % WT).

An SR protein can be rendered dispensable by increasing the complementarity of a mammalian splicing signal to a U snRNA. (A) Schematic diagram showing the sequence of the wild-type dsx-MS2 5′ splice site and the mutant derivative that increases complementarity to U6 snRNA [dsx-MS2(U6 mut)]. The sequence of the base-pairing region of U6 snRNA is shown; mutated nucleotides are indicated in bold. (B) In vitro splicing analysis of the wild-type dsx-MS2 pre-mRNA substrate and dsx-MS2(U6 mut) mutant derivative in the presence of the MS2–(ASF)RS fusion protein and/or the ASF/SF2 protein. Splicing reactions were performed in S100 extract.

The RS domain–splicing signal interaction promotes RNA base-pairing. (A) Psoralen cross-linking experiment. Reaction mixtures containing a uniformly ³²P-labeled dsx-MS2 pre-mRNA; unlabeled in vitro synthesized U2 snRNA; or U2 snRNAΔBPRS, the MS2–(ASF)RS fusion protein, and psoralen were irradiated with UV light to induce RNA–RNA cross-links and the RNA species were fractionated on a denaturing polyacrylamide gel. The cross-linked RNA product is indicated by the arrow. (B) The position of the U2 snRNA–RNA contact was mapped by RNase H protection assays. Uniformly labeled dsx-MS2 pre-mRNA mixed with unlabeled U2 snRNA (left) or the eluted RNA–RNA cross-linked band (right) was incubated with a ∼20-nt DNA oligonucleotide complementary to the branchpoint, Py-tract, 3′ splice site, or MS2-binding site, and subjected to RNase H treatment in the absence (left) or presence (right) of psoralen cross-linking. Shown below is a schematic diagram depicting the dsx-MS2 substrate, the positions of the DNA oligonucleotides, and the expected sizes of the larger of the two RNase H-generated fragments. The portion of the gel containing the larger of the two cleavage products is shown.

In vitro splicing of yeast actin 5′ splice site mutants promoted by an RS domain. (A) In vitro splicing analysis of the wild-type actin-MS2(I) pre-mRNA substrate and mutant derivatives (shown on the left), performed as described in Figure 1A. (B) MS2–(U2AF⁶⁵)RS and MS2–(U2AF³⁵)RS were analyzed for their ability to rescue splicing of the actin-MS2(I)/6U-C and actin-MS2(I)/5G-A mutant substrates. (C) UV cross-linking analysis of the RS domain–pre-mRNA interaction with the wild-type actin-MS2(I) (top), actin-MS2(E2)/6U-C (middle), and actin-MS2(E2)/1G-C (bottom) substrates, which were labeled with a single ³²P-G at the 5′ splice site (5′ss), MS2-binding site (MS2), branchpoint (BP), Py-tract (Py), 3′ splice site (3′ss), as shown in the schematic diagram (bottom).

Promotion of yeast in vivo splicing by an RS domain requires the SR protein kinase, Sky1p. (A) Immunoblot analysis of MS2–(ASF)RS phosphorylation in wild-type yeast and a sky1-Δ deletion strain. (B) In vivo splicing assays in wild-type or sky1-Δ deletion strains coexpressing MS2–(ASF)RS and a wild-type or mutant actin-MS2(E2) (left) or actin-MS2(I) pre-mRNA (right) derivative. The amount of spliced product was quantitated relative to the amount of product obtained with wild-type substrate in wild-type yeast (percent wild type: % WT).

An RS domain tethered to RNA selectively contacts a double-stranded RNA region in the absence of other proteins. (A, left) UV cross-linking analysis of the ability of the RS domain to selectively contact a double-stranded RNA region. The dsx-MS2 pre-mRNA substrate was site-specifically labeled at the branchpoint (indicated by the asterisk) and incubated with the MS2–(ASF)RS fusion protein in the presence or absence of a molar excess of a 15-nt RNA oligonucleotide complementary to the branchpoint region (BP-oligo indicated by a line). Splicing reactions were performed in the presence of HeLa nuclear extract (NE) or micrococcal nuclease-treated nuclear extract (MNase-NE). (Right) UV cross-linking analysis in the presence of NE or ATP-depleted NE (NEΔATP), and in the presence or absence of exogenously added ATP. (B) UV cross-linking analysis in the absence of exogenously added ATP, and in the presence of NE or ATP-depleted NE (NEΔATP). Immunoprecipitations were performed with an anti-Flag antibody or the mAb104 antibody, as shown. (C, top) Immunoblot analysis of MS2–(ASF)RS phosphorylation following expression of MS2–(ASF)RS in a bacterial strain coexpressing the RS domain protein kinase SRPK. (Bottom) UV cross-linking analysis of the ability of phosphorylated MS2–(ASF)RS to contact a double-stranded RNA in the absence of exogenously added ATP and in the presence and absence of nuclear extract. (D, top) UV cross-linking analysis of the ability of phosphorylated MS2–(ASF)RS to contact RNA in the presence of U2 snRNA or a mutant derivative lacking the branchpoint recognition sequence (U2 snRNAΔBPRS). (Bottom) Schematic diagram showing the sequences of the dsx-MS2 branchpoint (the branchpoint adenine is underlined), the base-pairing region of U2 snRNA, and the nucleotides deleted in the U2 snRNAΔBPRS mutant. (E) Analysis of the sequence, distance, and orientation dependence of the interaction between the RS domain and the double-stranded RNA region. RNA containing an MS2-binding site and lacking natural intron sequences was site-specifically labeled at ∼50, 100, or 150 nt either upstream (top) or downstream (bottom) from the MS2-binding site. Each reaction mixture contained an RNA oligonucleotide complementary to the site-specifically labeled region (indicated by a line), and the ability of the phosphorylated MS2–(ASF)RS to contact the double-stranded RNA region was analyzed by UV cross-linking analysis. (Bottom) Sequences of the RNA oligonucleotides.