The polypyrimidine tract binding protein (PTB) represses splicing of exon 6B from the beta-tropomyosin pre-mRNA by directly interfering with the binding of the U2AF65 subunit

Abstract

Splicing of exon 6B from the beta-tropomyosin pre-mRNA is repressed in nonmuscle cells and myoblasts by a complex array of intronic elements surrounding the exon. In this study, we analyzed the proteins that mediate splicing repression of exon 6B through binding to the upstream element. We identified the polypyrimidine tract binding protein (PTB) as a component of complexes isolated from myoblasts that assemble onto the branch point region and the pyrimidine tract. In vitro splicing assays and PTB knockdown experiments by RNA interference demonstrated that PTB acts as a repressor of splicing of exon 6B. Using psoralen experiments, we showed that PTB acts at an early stage of spliceosome assembly by preventing the binding of U2 snRNA on the branch point. Using UV cross-linking and immunoprecipitation experiments with site-specific labeled RNA in PTB-depleted nuclear extracts, we found that the decrease in PTB was correlated with an increase in U2AF65. In addition, competition experiments showed that PTB is able to displace the binding of U2AF65 on the polypyrimidine tract. Our results strongly support a model whereby PTB competes with U2AF65 for binding to the polypyrimidine tract.

FIG. 1.

PTB is a component of myoblast protein complexes assembled onto BP and PY. (A) Schematic representation of the two mutually exclusive exons showing the regulatory elements upstream of exon 6B. PmlI and SexAI are restriction enzymes used to construct the plasmid pPmac SexAI. The sequences of the RNAs used for affinity purification of protein complexes are shown below. The A in bold type indicates the branch point upstream of exon 6B (22). (B) Two-dimensional gel electrophoresis of proteins from QM7 myoblast nuclear extracts assembled onto BP, PY, and 3′6B RNAs. Proteins were revealed by silver staining. The insets shown to the right of the BP and PY RNAs represent parts of the Western blot analysis performed on a two-dimensional gel with polyclonal antibodies against PTB. (C) Western blot analysis with anti-PTB of protein complexes assembled on wild-type or mutant RNAs. BPmtPTB and PYmtPTB are mutant BP and PY RNAs, respectively, in which putative PTB binding sites have been mutated (Fig. 1A). Odd and even lanes correspond to QM7 nuclear extracts from myoblasts (mb) and myotubes (mt) assembled onto RNAs, respectively.

FIG. 2.

PTB represses in vitro splicing of exon 6B. (A) Schematic representation of the pre-mRNA used in in vitro splicing experiments. The X mark indicates the mutation of the 5′ splice site of exon 6A. The deletion of the intronic S4 enhancer sequence is indicated. (B) Addition of increasing amounts of Cs8 RNA competitor activates splicing of exon 6B, and adding back of recombinant PTB1 restores splicing inhibition. Labeled pre-mRNA 5-K6AΔ4-6B7 was incubated in 40% HeLa nuclear extracts for 90 min at 30°C in the absence (lane 2) or in the presence of 2, 4, 5, 6, and 8 pmol of Cs8 (lanes 3 to 8, respectively). Lane 1, pre-mRNA alone. In lanes 9 to 13, 4 pmol of Cs8 was added in the absence (lane 9) or in the presence of 0.25, 0.5, 0.75, or 1 μg of recombinant PTB (lanes 10 to 13, respectively). The products of the splicing reactions were separated by denaturing 5% PAGE. The identities of the splicing products are diagrammed to the left of the gel. (C) Immunodepletion of PTB from nuclear extracts activates splicing of exon 6B. Splicing experiments were carried out under the same conditions as for panel B, with mock-depleted (lanes 1 to 3) and PTB-immunodepleted (lanes 4 to 9) nuclear extracts in the presence of 0.2, 0.3, and 0.6 μg of recombinant PTB1 (lanes 7 to 9) for 60 min (lanes 1 and 4) or 90 min (lanes 2, 3, and 5 to 9). Note that the incubation of nuclear extracts with beads alone (mock-depleted nuclear extracts) resulted in a modification of the ratio between the splicing of exon 6A over the splicing of exon 6B (in panels B and C, compare lanes 2). We have no clear explanation for this finding except that the incubation of nuclear extracts, which decreases the splicing efficiency of the two splicing pathways, could allow a better competition between the splicing of exon 5 to exon 6B over the splicing of exon 5 to exon 6A.

FIG. 3.

The 3′ region upstream of exon 6B makes pre-mRNA responsive to PTB. (A) Schematic representation of the Glo-βTm pre-mRNA used in in vitro splicing experiments. The intronic region upstream of exon 6B containing the BP, PY, and 3′ 6B fragments is shown. (B) In vitro splicing of the Glo-βTm pre-mRNA. Splicing experiments were performed as in Fig. 2B, with mock-depleted extracts (lane 1) and PTB-immunodepleted nuclear extracts (lanes 2 to 5) in the absence (lane 2) or in the presence of 0.05, 0.1, and 0.2 μg of recombinant PTB1 (lanes 3 to 5, respectively) for 90 min.

FIG. 4.

PTB knockdown activates splicing of exon 6B in HeLa cells. (A) Western blot analysis of PTB knockdown using anti-PTB antibodies. Twenty micrograms of total protein was loaded per lane. Lanes 1 and 5, cells not treated; lanes 2 to 4, cells treated with 20, 60, or 180 pmol PTB siRNA, respectively; lanes 6 to 8, cells treated with 20, 60, or 180 pmol luciferase siRNA, respectively. The blot was probed again with an anti-hnRNP A1 monoclonal antibody. (B) Schematic representation of the wild-type 980 βTm reporter minigene. The location of primers used for RT-PCR is indicated. Schematic diagrams of PCR-amplified splicing products are shown below. (C) HeLa cells treated with 180 pmol siRNA targeting PTB (lanes 1 to 5), with the same amount of a control luciferase siRNA (lanes 6 to 10) or without siRNA (lane 11), were transfected with 0.5 μg of the wild-type 980 minigene and with increasing amounts of expression vector coding for PTB1 (10, 30, 100, or 300 ng in lanes 2 to 5 and lanes 7 to 10). The splicing products were analyzed by RT-PCR and quantified with a PhosphorImager. (D) Wild-type and mutant minigenes were cotransfected with 180 pmol PTB siRNA or Luc siRNA in HeLa cells and analyzed by RT-PCR as for panel C. The percentage of exon 6B inclusion is shown as the mean ± standard deviation for at least four experiments.

FIG. 5.

Depletion of PTB from nuclear extracts promotes the interaction of U2 snRNP. (A) Psoralen cross-linking reactions using ³²P-labeled pPmac-sexA1 were carried out with PTB-depleted (lanes 1 to 5) or mock-depleted (lanes 6 to 10) nuclear extracts in the presence of 50 ng (lanes 3 and 8), 100 ng (lanes 4 and 9), or 150 ng (lanes 5 and 10) of recombinant PTB1. RNAs were recovered and analyzed on a 5% polyacrylamide denaturing gel. Cross-linked species are indicated on the left. (B) Identification of cross-linked products. ³²P-labeled pPmac-sexA1 was incubated with PTB-depleted nuclear extracts. Purified RNA from cross-linked reactions was mock-treated (lane 1) or treated with RNase H and oligonucleotides complementary to snRNA, as indicated at the top of each lane. Lanes 2 and 3, 20 and 200 pmol oligonucleotide complementary to nucleotides 29 to 42 of U2 snRNA, respectively; lanes 4 and 5, 10 and 100 pmol oligonucleotide complementary to nucleotides 1 to 15 of U2 snRNA, respectively; lanes 6 and 7, 10 and 100 pmol oligonucleotide complementary to nucleotides 66 to 77 of U1 snRNA, respectively. (C) Mapping of cross-linking sites on the RNA. Primer extension of U2 and U1-BP gel-purified cross-linked species was performed with an oligonucleotide complementary to the intron between exons 6A and 6B. Lanes 1 and 2, primer extension from U2 and U1-BP cross-linked species, respectively; lane 3 (Ex), primer extension from pPmac-SexA1; lanes 4 to 7, dideoxy sequencing reaction performed on pPmac-SexA1. Black and white arrows indicate transcription stops on pPmac-SexA1 with U2 and U1-BP cross-linked species, respectively. The positions of psoralen cross-links are indicated by circles on the sequence of pPmac-SexA1, shown on the right. (D) The cross-linked sites on U1-BP snRNA were mapped with an oligonucleotide complementary to nt 66 to 77 of U1 snRNA on the purified U1-BP species. Lane 1, primer extension of U1-BP snRNA cross-linked species; lanes 2 and 3, primer extension of snRNA U1 from HeLa nuclear extracts; lane 4 (Ex), primer extension of snRNA U1 transcribed from pBSU1 (25); lanes 5 to 8, dideoxy sequencing of snRNA U1 transcribed from pBSU1. Black arrows indicate specific transcription stops on U1 snRNA. The positions of the psoralen cross-links are indicated by circles on the sequence of U1 snRNA, shown on the right.

FIG. 6.

(A) The interaction of U2 snRNP on the branch point requires the presence of ATP. Psoralen cross-linking experiments were performed as described in the legend to Fig. 5A, with or without ATP (lanes 1 and 2, respectively). The bracket on the right indicates intramolecular cross-linked species. (B) The interaction of U2 snRNP on the branch point requires a downstream 5′ splice site. Psoralen cross-linking experiments were performed as for panel A with wild-type pPmac-SexA1 (lanes 1 to 4) or with 5′ splice site mutated RNA (lanes 5 to 8) or with deleted 5′ splice site (lanes 9 to 12) in the absence (lanes, 1, 2, 5, 6, 9, and 10) or in the presence (lanes, 3, 4, 7, 8, 11 and 12) of 0.5 pmol of Cs8/μl of nuclear extracts. For each RNA, odd and even lanes are duplicated samples. The star represents intramolecular cross-linked species. (C) PTB prevents spliceosome complex formation. The pre-mRNA containing exon 6A, intron, and exon 6B was incubated under splicing conditions without (lanes 1 to 5) or with (lanes 6 to 10) 0.5 pmol of Cs8/μl of nuclear extracts for 0 min (lanes 1 and 6), 15 min (lanes 2 and 7), 30 min (lanes 3 and 8), 60 min (lanes 4 and 9), and 90 min (lanes 5 and 10) at 30°C. H and SP represent heterogeneous and spliceosome complexes, respectively. Lanes 11 to 15, control pre-mRNAs.

FIG. 7.

PTB prevents the binding of U2AF65. (A) Western blot analysis of mock-depleted (M) HeLa nuclear extracts or extracts depleted with the Cs8 RNA (ΔR) or immunodepleted with a commercial monoclonal PTB antibody (Δα). The same blot was probed with the monoclonal anti-PTB antibody, then with the monoclonal anti-U2AF65 antibody. Note that the doublet below the U2AF65 signal corresponds to the remaining PTB signal. (B) Uniformly labeled pPmac-SexA1 RNA was incubated in mock-depleted (M) or PTB-depleted (ΔR and Δα) nuclear extracts under splicing conditions for 15 min at 30°C in the absence (lanes 1, 3, 5, and 7) or in the presence (lanes 2, 4, 6 and 8) of 1.2 μg of recombinant PTB1. After UV cross-linking, the samples were separated by SDS-PAGE (lanes 1 to 8, X-link) or immunoprecipitated with monoclonal anti U2AF65 antibodies (lanes 9 to 16, αU2AF65). The amount of UV cross-linked samples (lanes X-link) that were subjected to SDS-PAGE represents 1/10 of the volume of the immunoprecipitated samples. The positions of PTB and U2AF65 are indicated. (C) Uniformly labeled pPmac-SexA1 was incubated with 0.5, 1, 2, 4.3, 8.7, and 17.4 pmol of GST-U2AF65 (lanes 1 to 6, respectively) or with 0.44, 0.88, 1.76, 3.5, 7, and 14 pmol of his-PTB1 (lanes 7 to 12, respectively). (D) Uniformly labeled pPmac-SexA1 was incubated with 9 pmol of recombinant GST-U2AF65 (lanes 1 to 12) and with 0.22 pmol (lanes 2 and 8), 0.44 pmol (lanes 3 and 9), 0.88 pmol (lanes 4 and 10), 1.76 pmol (lanes 5 and 11), and 3.5 pmol (lanes 6 and 12) of recombinant PTB1 in the absence (lanes 1 to 6) or in the presence (lanes 7 to 12) of immunodepleted HeLa nuclear extracts. The samples were UV cross-linked, and the proteins were separated by SDS-PAGE.

FIG. 8.

PTB competes with U2AF65 for binding. (A) A truncated part of pPmac-SexA1 is shown with the intron upstream of exon 6B indicated in lowercase letters and the beginning of exon 6B in capital letters. The positions of site-specific labeling are indicated. The branch point is shown in bold. The positions of the different RNA fragments are indicated with their lengths. (B) U2AF65 interacts mainly with BP-PY. Gel shift experiments were performed without (lanes 1) or with (lanes 2 to 5, respectively) 1.7 nM, 5.4 nM, 17 nM, and 54 nM concentrations of recombinant GST-U2AF65. The K_d(app) was estimated by the concentration of protein that gives 50% of the RNA bound to the protein. (C) Binding of PTB to the intronic sequences upstream of exon 6B. Gel retardation assays were performed as for panel B without (lanes 1) or with (lanes 2 to 5, respectively) 2.8, 8.8, 28, and 88 nM recombinant his-PTB1. Note that higher-order complexes were detected at high PTB concentrations, as observed by others (2, 23). (D) UV cross-linking experiments on site-specifically labeled pPmac-SexA1. The reactions were performed as described for Fig. 7B with mock-depleted (M) or PTB-depleted (ΔR and Δα) nuclear extracts. After UV cross-linking, the samples were separated by SDS-PAGE (lanes 1 to 4, 9 to 12, and 17 to 20 [X-link]) or immunoprecipitated with anti U2AF65 antibodies (lanes 5 to 8, 13 to 16, and 21 to 24 [αU2AF65]). X denotes unknown cross-linked proteins with the RNA labeled at position 34 downstream of the branch point. The lower panel shows same samples as the upper panel, but for each position, different times of exposure on the PhosphorImager screen are presented. (E) Immunoprecipitation with anti-PTB antibodies. UV cross-linking experiments were performed under the same conditions as described for panel D and immunoprecipitated with anti-PTB antibodies.