A functional interaction between the carboxy-terminal domain of RNA polymerase II and pre-mRNA splicing

Abstract

In the preceding study we found that Sm snRNPs and SerArg (SR) family proteins co-immunoprecipitate with Pol II molecules containing a hyperphosphorylated CTD (Kim et al., 1997). The association between Pol IIo and splicing factors is maintained in the absence of pre-mRNA, and the polymerase need not be transcriptionally engaged (Kim et al., 1997). The latter findings led us to hypothesize that a phosphorylated form of the CTD interacts with pre-mRNA splicing components in vivo. To test this idea, a nested set of CTD-derived proteins was assayed for the ability to alter the nuclear distribution of splicing factors, and to interfere with splicing in vivo. Proteins containing heptapeptides 1-52 (CTD52), 1-32 (CTD32), 1-26 (CTD26), 1-13 (CTD13), 1-6 (CTD6), 1-3 (CTD3), or 1 (CTD1) were expressed in mammalian cells. The CTD-derived proteins become phosphorylated in vivo, and accumulate in the nucleus even though they lack a conventional nuclear localization signal. CTD52 induces a selective reorganization of splicing factors from discrete nuclear domains to the diffuse nucleoplasm, and significantly, it blocks the accumulation of spliced, but not unspliced, human beta-globin transcripts. The extent of splicing factor disruption, and the degree of inhibition of splicing, are proportional to the number of heptapeptides added to the protein. The above results indicate a functional interaction between Pol II's CTD and pre-mRNA splicing.

Figure 1

Fusion proteins derived from Pol II's CTD. The largest subunit of RNA Polymerase II (Pol II LS) is illustrated schematically (top). An expanded view of the CTD shows 52 heptapeptide repeats represented by variably shaded boxes. Lightly shaded boxes represent consensus repeats (YSPTSPS) and more darkly shaded boxes represent variant repeats (see Corden et al., 1985; Wintzerith et al., 1992). The CTD coding sequence was unidirectionally truncated from the COOH terminus and recombinantly fused to the Flag^® peptide (Flag symbol) or βGalactosidase (oval symbol). The resulting fusion proteins are described by nomenclature that begins with the NH₂ terminus and ends with the COOH terminus, including the number of heptapeptide repeats. Symbols are summarized in the key (for details see Materials and Methods).

Figure 7

Plasmids expressing human β-globin transcripts and CTD-derived fusion proteins. (A) A wildtype human β-globin gene with a downstream SV40 enhancer (SV40E) was inserted into an EcoRV site in multiple plasmids that express Flag-tagged proteins or βGal (Fig. 1). For brevity the illustration depicts the insertion of various protein-encoding sequences into a site upstream of the β-globin gene. The Flag-tagged proteins and βGal coding sequences are under the control of the CMV promoter (CMVp) (for details see Materials and Methods). β-Globin introns are represented by open boxes, exons by black boxes, and noncoding flanking sequences open boxes at the ends of the gene. β-Globin and CMV promoters are indicated by bent arrows. The resulting constructs are generically termed “FusionProteinβ-globin [+].” The plus sign indicates that the two genes are oriented in the same direction. The primers (P1 and P2) hybridize with complementary (cDNA) sequences within exons 1 and 2, respectively. PCR amplification with P1 and P2 yields 170-nt and 300-nt DNA fragments corresponding to spliced and unspliced transcripts, respectively. The 343-nt RNA probe used for RNase protection is shown below the β-globin gene. The open box on this probe represents a nonhybridizing portion derived from pBluescript, and the black bar hybridizes with a 276-nt segment of the unspliced β-globin transcript. The 276-nt segment spans an intronexon boundary including 203 nucleotides of exon 2 and 73 nucleotides of intron 1. Therefore, the spliced and unspliced β-globin transcripts protect 203 and 276 nucleotide segments of the probe, respectively. (B) A wild-type human β-globin gene with a downstream SV40 enhancer (SV40E) was also inserted in the opposite orientation of the EcoRV site in the plasmids expressing Flag-tagged proteins or βGal. The resulting constructs are generically termed “FusionProteinβ-globin [−].” The minus sign indicates that the two genes are oriented in the opposite direction. For convenience, the protein-encoding sequences are not shown. (C) A thalassemic human β-globin gene with a downstream SV40 enhancer (SV40E) was inserted in the positive orientation into the EcoRV site in the plasmids expressing Flag-tagged proteins or βGal. The resulting constructs are generically termed “FusionProteinβ-globin^thal [+].” The thalassemic allele is mutated at first residue of intron 1 (G to A transition) (delta symbol). Splicing of exons 1 and 2 is achieved by using three cryptic 5′ splice sites and the normal 3′ splice site (see Caceres et al., 1995). The oligonucleotide used for RNase protection spans the 3′ splice site, but it is downstream of the cryptic 5′ splice sites. Therefore, all three variably spliced transcripts register as 203 nucleotide RNAs in the RNase protection assay. For convenience, the protein-encoding sequences are not shown.

Figure 2

Expression, in vivo phosphorylation, and nuclear localization of CTD-derived fusion proteins. (A and B) Immunoblotting. CV1 cells were transfected with each of the plasmids listed above the panels (see Materials and Methods). 2 d later, the cells were lysed in SDS sample buffer, subjected to 5–15% gradient SDS-PAGE, and immunoblotted with the antibodies listed below each panel. mAb M2 is directed against the Flag^® epitope, anti-βGal is directed against βGalactosidase, and mAbs H5 and H14 are directed against CTD phosphoepitopes (Kim et al., 1997). Numbers at the margins indicate apparent molecular weights in kilodaltons. pSF, control plasmid expressing a Flag^® tagged ∼30-kD segment of human β-spectrin. IIo, hyperphosphorylated largest subunit of Pol II. (C) Immunofluorescence microscopy. CV1 cells were transfected with pF-CTD52. 2 d later the cells were fixed and double immunostained with anti-Flag^® mAb M2 and anti-CTD mAb H14 (see Materials and Methods). Anti-Flag staining is visualized by rhodamine (left panel) and mAb H14 staining is visualized by FITC (right panel). The cell at the top expresses the F-CTD52 protein, and the cell at the bottom is an untransfected control. Note that mAb M2 labeling is almost exclusively intranuclear. In addition, the untransfected cell nucleus is weakly immunostained by mAb H14, whereas the transfected cell nucleus is intensely labeled. mAbs M2 and H14 stain the diffuse nucleoplasm, but they also stain ∼50 discrete “dots.” Bar, 10 μM.

Figure 4

Addition of heptapeptide repeats to Flag-tagged fusion proteins potentiates their disruptive effect on B1C8 speckles. CV1 cells were transfected with plasmids encoding fusion proteins listed at the left margin. 2 d later, the cells were fixed and double immunostained as described in Fig. 3. The CTD-derived fusion proteins were immunolocalized with anti-Flag^® mAb M2 (A, D, G, J, and M). A 160-kD SR-related family splicing factor was immunolocalized with mAb B1C8 (B, E, H, K, and N). Digital images were pseudocolored and merged as described in Fig. 3 (C, F, I, L, and O). Red pseudocolor indicates distribution of CTD-derived fusion proteins. Green pseudocolor indicates distribution of SR splicing factor B1C8. Yellow pseudocolor indicates overlap between red and green. White dots, nuclei expressing fusion proteins; thick arrows, intact B1C8 speckles; thin arrows, Flag-tagged CTD-derived protein in discrete nuclear sites; arrowhead, B1C8 speckles immediately adjacent to F-CTD1 (B and C) and FCTD6 (H and I). Bar, 10 μm.

Figure 3

CTD52 disrupts the speckled distribution of SR splicing factor B1C8. CV1 cells were transfected with plasmids encoding fusion proteins listed at the left margin. 2 d later, the cells were fixed and double immunostained (see Materials and Methods). CTD-derived fusion proteins and control proteins were immunolocalized with anti-Flag^® mAb M2 (A, D, and G) or anti-βGal (J and M). A 160-kD SR-related family splicing factor (Blencowe et al., 1995) was immunolocalized with mAb B1C8 (B, E, K, and N). ND55 was immunolocalized with mAb 138 (Ascoli and Maul, 1991). Red pseudocolor indicates distribution of Flag-tagged or βgal-linked fusion proteins. Green pseudocolor indicates distribution of endogenous nuclear proteins B1C8 or ND55. Red and green digital images were merged (C, F, I, L, and O), and areas of overlap between the distributions of transiently expressed fusion proteins and endogenous proteins are pseudocolored yellow. White dots, nuclei expressing fusion proteins; single arrows, B1C8 speckles; double arrows, ND55 in N10/PML domains (H). Bars, 10 μm.

Figure 5

Relationship between CTD length and disruptive effect on B1C8 speckles. CV1 cells were transfected with plasmids encoding the fusion proteins listed above the histogram bars. 2 d later, the cells were fixed and double stained with antibodies directed at the Flag^® epitope and B1C8 as described in Fig. 4. The pattern of B1C8 staining in each transfected cell nucleus was scored as “intact” (20–50 prominent speckles) or “disrupted” (diffuse pattern or diminutive speckles). Data were pooled from multiple experiments performed on different days. 150–250 nuclei were scored for each plasmid.

Figure 6

CTD52 disrupts the speckled distribution of Sm snRNPs without altering the distribution of p80 coilin. Transfection and immunostaining were performed as described above. (A– C) CV1 cells were transfected with pF-CTD52, fixed and double immunostained with mY12 (directed at Sm snRNPs; Lerner et al., 1981), and mAb H5 (directed at CTD phosphoepitopes; Bregman et al., 1995; Kim et al., 1997). A nucleus expressing F-CTD52 is identified by the intense immunostaining with mAb H5 (upper right corner). Three untransfected cell nuclei are identified by weaker mAb H5 immunostaining. (D–F) CV1 cells were transfected with pβGal-CTD52. A nucleus expressing βGal-CTD52 is identified by intense immunostaining with mAb anti-βGal, and three untransfected cell nuclei are identified by faint immunostaining with mAb anti-βGal. (G–I) CV1 cells were transfected with pF-CTD13. The distribution of F-CTD13 is revealed by red pseudocolor in three transfected cell nuclei (G and I). The distribution of p80 coilin is revealed by green pseudocolor in the three cells expressing F-CTD13, and in one untransfected cell (H and I). White dots, nuclei expressing CTD-derived fusion proteins; single arrows, speckles stained by mAb Y12; double arrows, p80 coilin in coiled bodies (H and I). Bars, 10 μm.

Figure 8

CTD-derived fusion proteins block the accumulation of spliced, but not unspliced, β-globin transcripts in vivo. (A) HeLa cells were transfected with pF-CTDless.1β-globin [+], pFCTD₁₃ β-globin [+], or pF-CTD₅₂ β-globin [+]. 1 d later, total RNA was prepared from the cells as described in the Materials and Methods. The RNA was reverse transcribed and amplified by PCR using primers P1 and P2 (see Fig. 7 A). The PCR products were separated by agarose gel electrophoresis, stained with ethidium bromide, and photographed. (B) HeLa cells were transfected with pF-CTDless.1β-globin [+], pFCTD₁₃ β-globin [+], pF-CTD₅₂ β-globin [+], pF-CTDless.1 β-globin [−], pFCTD₁₃ β-globin [−], or pF-CTD₅₂ β-globin [−]. 1 d later, RNA was prepared and subjected to the RNase protection assay described in Fig. 7 and Materials and Methods. Protected RNAs were separated by electrophoresis and processed for autoradiography. (C) HeLa cells were transfected with pF-CTDless.1 β-globin^thal [+], pF-CTD₁₃ β-globin^thal [+], or pF-CTD₅₂ β-globin^thal [+]. 1 d later, RNase protection assays were performed as described in B. (D) HeLa cells were transfected with pF-CTDless.3βglobin^thal [+], pβGalβ-globin^thal [+], pFCTDless.1β-globin^thal [+], pF-CTD₁ β-globin^thal [+], pF-CTD₆ β-globin^thal [+], pFCTD₁₃ β-globin^thal [+], and pF-CTD₅₂ β-globin^thal [+]. 1 d later, RNase protection assays were performed as described in B. In A–D, the expressed Flag-tagged proteins are indicated at the top of the panel. The β-globin transcripts, and their orientation relative to the CMV driven transcription unit, are indicated below the panels. U, unspliced; S, spliced; P, intact probe; C, control yeast RNA; R, RNase added; M, molecular weight markers. MWs are indicated in base pairs at the left hand margin.