Phosphorylation switches the general splicing repressor SRp38 to a sequence-specific activator

Abstract

SRp38 is an atypical SR protein that functions as a general splicing repressor when dephosphorylated. We now show that phosphorylated SRp38 functions as a sequence-specific splicing activator. Unlike characterized splicing activators, SRp38 functions in the absence of other SR proteins but requires a cofactor for activity. SRp38 was able to induce formation of splicing complex A in the absence of the cofactor, but this factor was necessary for progression to complexes B and C. Mechanistically, SRp38 strengthens the ability of the U1 and U2 small nuclear ribonucleoproteins to stably recognize the pre-mRNA. Extending these findings, analysis of alternative splicing of pre-mRNA encoding the glutamate receptor B revealed that SRp38 alters its splicing pattern in a sequence-specific manner. Together, our data demonstrate that SRp38, in addition to its role as a splicing repressor, can function as an unusual sequence-specific splicing activator.

Figure 1

SRp38 functions as a sequence-specific activator of splicing. (a) Schematic representation of β-globin and its derivatives, containing either three copies of the SRp38 consensus sequence or a control sequence. (b) Activation of β-SRp38 pre-mRNA splicing by SRp38. Splicing was performed in S100 supplemented with 50 ng (lanes 3, 5, 9 and 11) or 100 ng (lanes 4, 6, 10 and 12) of baculovirus-produced His-SRp38, with or without a nuclear fraction (NF40–60), as indicated above. Products of splicing were analyzed by denaturing PAGE and autoradiography. Splicing products are indicated schematically. (c) Dose-dependent splicing activation of SRp38. The indicated amounts of His-SRp38 were added to splicing reactions performed in S100 supplemented with NF40–60.

Figure 2

Characteristics of SRp38-dependent splicing activation. (a) RS domain requirements for activation. β-SRp38 pre-mRNA was incubated in S100 activated by 50 ng of GST-SRp38, GST–SRp38-2 or GST–SRp38 RBD, or 8 ng of dephosphorylated His-SRp38 (dSRp38), respectively, in the absence (lanes 2, 4, 6 and 8) or presence (lanes 3, 5, 7 and 9) of NF40–60. (b) Repression of splicing by dSRp38. His-dSRp38 (0 ng, 2 ng, 4 ng and 20 ng) was incubated with S100 plus phosphorylated His-SRp38 (20 ng) and NF 40–60 (lanes 10–12). (c) Role of the RBD in SRp38-dependent activation. The indicated amounts of purified GST–SRp38 RBD, His–SC35 RBD or GST–hnRNP G RBD were added to SRp38-dependent splicing assays as indicated. (d) SR proteins do not cooperate with SRp38 for splicing activation. Increasing amounts of purified SR proteins were added to reactions performed in S100 alone (lanes 2–4) or in the presence of 50 ng of GST-SRp38 (lanes 5–7). (e) NF40–60 coactivator activity is specific for SRp38. His-SRp38, His-ASF or His-SC35 (50 ng) was incubated with β-SRp38 or β-SRp38 RNAs in S100 alone or supplemented with NF40–60. (f) Splicing stimulation by ASF/SF2 or SC35 at high concentrations. His-tagged SRp38, ASF/SF2 and SC35 (300 ng) was incubated with β-SRp38 RNA in S100.

Figure 3

SRp38 promotes formation of spliceosomal complex A. (a) Spliceosome-assembly assays were carried out in S100 complemented with the indicated components and the β-SRp38 pre-mRNA. Splicing complexes were resolved on a 1.5% low-melting agarose gel. (b) SRp38 activates AdML-SRp38 in in vitro splicing assays. Splicing was performed in S100 supplemented with 50 ng of GST-tagged SRp38 with (lane 1) or without (lane 3) NF40–60. Products of splicing were analyzed by denaturing PAGE and autoradiography. (c) Spliceosome assembly was performed the same as in a, except with the AdML-SRp38 pre-mRNA. NE, nuclear extract. (d) Spliceosome-assembly assays were carried out as in c, except with the additions indicated at the top. Anti-U1, anti-U2, anti-U5 or anti-U6 snRNA oligonucleotides (5 µM) were added to reaction mixtures. Endogenous ATP was depleted by preincubating S100 at 30 °C for 40 min.

Figure 4

SRp38 interacts with both U1 and U2 snRNP complexes on the SRp38 substrate. (a,b) Purified U1 snRNP (a) or U2 snRNP (b), GST-tagged SRp38 (150 ng) or GST-tagged SRp38 RBD (250 ng) were incubated with β-SRp38 pre-mRNA (lanes 1–9) or β-control pre-mRNA (lanes 10–14). Following addition of heparin (0.8 mg ml⁻¹), complexes were resolved by 5% nondenaturing PAGE. Increasing amounts (300 ng and 600 ng) of U1 snRNP (a) or U2 snRNP (b) were incubated with SRp38 and β-SRp38 RNA (lanes 4, 5 and 6). The complexes formed are indicated with brackets. GST-tagged SRp38 RBD (250 ng) was incubated with β-SRp38 RNA (lane 6), and increasing amounts of U1 or U2 snRNPs (150 ng, 300 ng or 600 ng) were added (lanes 7–9). SRp38 (150 ng) was incubated with β-control pre-mRNA alone (lane 13) or β-control pre-mRNA plus U1 or U2 snRNPs (300 ng; lane 14). (c) His-tagged ASF/SF2 was incubated with β-SRp38 in the absence (lane 2) or in the presence of U1 snRNP (lane 3) or U2 snRNP (lane 4). (d) Splicing reactions were performed with the β-SRp38 RNA. The (5′ ss-Py) oligonucleotide was preincubated with nuclear extract (NE) or with S100 plus NF40–60 and SRp38 before splicing reactions were carried out. Final concentrations of the oligonucleotide are 0 µM, 0.1 µM , 0.5 µM, 1 µM and 2 µM. Note that 300 ng ASF/SF2 was used in the splicing reactions to achieve similar splicing levels as in the NE.

Figure 5

SRp38 binds to and activates splicing of RNA substrates containing the GluR-B Flip or Flop exon. (a) Comparison of a putative SRp38 binding motif in Flip and Flop exons from the murine GluR-B transcript with the SRp38 consensus recognition sequence. (b) Flop and Flip exons of the GluR-B pre-mRNA bind SRp38 specifically. Gel-shift assays with the indicated radiolabeled RNAs were performed with increasing amounts of GST-SRp38 (25 ng, 75 ng and 225 ng). Complexes were resolved by nondenaturing PAGE. (c) Flop and Flip exons of the GluR-B pre-mRNA do not bind His-SC35. Increasing amounts of purified His-SC35 (25 ng, 75 ng and 225 ng) were added to gel-shift assays containing the identical RNAs. (d) SRp38 activates splicing of RNA substrates containing Flip or Flop exons. The downstream β-globin exon was replaced with Flip or Flop exons, and splicing reactions were performed in S100 with increasing amounts of GST-SRp38 (20 ng, 40 ng and 100 ng) in the presence or absence of NF40–60 as indicated.

Figure 6

Loss of SRp38 promotes inclusion of the Flop exon in vivo. (a) Diagram of reporter plasmids containing mouse GluR-B truncated genomic sequences and three alternatively spliced products. The pair of primers used in the RT-PCR in b are shown as two reverse arrows. The RNA probe used in RNase protection assay in c is indicated above the three products, and the length of protected probe is on the side. (b) Comparison of alternatively spliced products from SRp38(+/+) and SRp38(−/−) DT40 cells. RT-PCR was performed with RNAs extracted from stably transfected DT40 cells and analyzed directly on a 1.5% agarose gel (lanes 1 and 2), or digested with StuI followed by agarose gel analysis (lanes 3 and 4). (c) RNase protection assay. RNA was extracted from stably transfected DT40 cells in the background of SRp38(+/+), SRp38(−/−) and SRp38(−/−) cells expressing exogenous hemagglutinin (HA)-SRp38. RNase protection assay was performed using the radiolabeled RNA probe indicated in a. Products were resolved by 6% denaturing PAGE.

Figure 7

Model for SRp38-dependent splicing activation. SRp38 binds the SRp38-dependent ESE in target transcripts and facilitates association of U1 and U2 snRNPs with the pre-mRNA to stabilize 5′-splice-site and branch-site recognition by interacting with U1 and U2 snRNPs, respectively. However, the spliceosomal A complex formed is stalled and requires an SRp38-specific cofactor to proceed through the splicing pathway.