Recognition of a functional 3′ splice site in pre-mRNA splicing requires a heterodimer of the proteins U2AF65/U2AF35. U2AF65 binds to RNA at the polypyrimidine tract1,2, whereas U2AF35 is thought to interact through its arginine/serine-rich (RS) domain with other RS-domain-containing factors bound at the 5′ splice site, assembled in splicing enhancer complexes, or associated with the U4/U6.U5 small nuclear ribonucleoprotein complex3,4,5,6,7. It is unclear, however, how such network interactions can all be established through the small RS domain in U2AF35. Here we describe the function of a U2AF35-related protein (Urp), which is the human homologue of a mouse imprinted gene. Nuclear extracts depleted of Urp are defective in splicing, but activity can be restored by addition of recombinant Urp. U2AF35 could not replace Urp in complementation, indicating that their functions do not overlap. Co-immunodepletion showed that Urp is associated with the U2AF65/U2AF35 heterodimer. Binding studies revealed that Urp specifically interacts with U2AF65 through a U2AF35-homologous region and with SR proteins (a large family of RS-domain-containing proteins) through its RS domain. Therefore, Urp and U2AF35 may independently position RS-domain-containing factors within spliceosomes.
We previously purified and cloned a mammalian serine kinase, SRPK1, which is highly specific for the RS-rich superfamily of splicing factors8,9,10. This stringent substrate specificity enabled us to do a yeast two-hybrid screen to search for new RS-domain-containing proteins. We isolated a positive clone from a HeLa complementary DNA library that contains a typical RS domain at the carboxy terminus (Fig. 1a). The recombinant protein was efficiently phosphorylated in the RS domain, like other RS-domain proteins, by SRPK1 in vitro (data not shown). This protein of 483 amino acids does not contain any RNA-recognition motif characteristic of the SR family of splicing factors11,12; instead, it contains two stretches of sequences, H1 and H2 (boxed in Fig. 1b), homologous to the human splicing factor U2AF35. We therefore named it Urp, for U2AF35-related protein. Northern blotting analysis indicates that both U2AF35 and Urp are ubiquitously expressed in all human tissues and cell lines examined (data not shown). Homology searches revealed that Urp had been independently cloned from humans as U2AF1-RS2 (ref. 13), which is nearly identical (94% amino-acid identity) to another human gene, U2AF1-RS1 (ref. 13) (Fig. 1b). These two human genes are equally homologous (∼80% identical) to two mouse genes14, one of which is imprinted15, and the other has a high transmission distortion in interspecific backcross progeny16. Thus, several U2AF35-related proteins are expressed in mammalian cells and may be members of the superfamily of RS-domain-containing splicing factors11.
To determine the function of Urp, we raised rabbit polyclonal antibodies against recombinant Urp expressed in bacteria. Immunoblotting of HeLa cell nuclear extracts demonstrated that anti-Urp antibodies specifically detect a single band of relative molecular mass 66K (Fig. 2a), which corresponds to endogeneous Urp as it matches the size of Urp translated in reticulocyte lysate (data not shown). To see whether Urp is essential for splicing, we constructed affinity columns with anti-Urp antibodies or preimmune serum. Control depletion with preimmune serum had no effect, but passing nuclear extract through an anti-Urp affinity column efficiently depleted it of Urp, as evidenced by western blotting and complete inhibition of splicing (data not shown). Although this result is consistent with Urp being a splicing factor, this Urp-depleted extract could not be complemented by recombinant Urp, probably because of co-depletion of other essential splicing factors (see below). We therefore did the immunodepletion in the presence of high salt (0.5 M KCl) to minimize co-depletion. As shown in Fig. 2c, human β-globin pre-mNA splicing in Urp-depleted nuclear extract was significantly decreased (compare lanes 1 and 2). The second step of the splicing reaction, which generates released lariat intermediate and ligated exons, was preferentially affected by Urp depletion under these conditions, indicating that Urp may play a role in this step; however, we could not rule out the possibility that it was required for the first step as we did not detect any accumulation of first-step splicing intermediates. To confirm the essential function of Urp in splicing, we restored Urp activity to a Urp-depleted nuclear extract with a glutathione-S-transferase (GST)–Urp fusion protein produced in baculovirus. Because Urp was insoluble in baculovirus-infected cells, we purified recombinant Urp from inclusion bodies, then isolated the protein by SDS–PAGE (Fig.2b). After denaturation and renaturation in the presence of bovine serum albumin, recombinant Urp restored splicing activity to the Urp-depleted nuclear extract in a concentration-dependent manner (Fig. 2c, lanes 4 to 6). There was no complementing activity associated with BSA (Fig. 2c, lane 3). We conclude that Urp is an essential splicing factor.
SR proteins have redundant functions, at least in vitro (for review, see refs 11, 12). U2AF35 was originally shown to be non-essential forconstitutive splicing because U2AF65 alone was sufficient to reconstitute a splicing-deficient nuclear extract from which the U2AF heterodimer was removed by poly(U) in the presence of high salt17. However, U2AF35 is essential for constitutive and enhancer-dependent splicing by immunodepletion and reconstitution experiments5. Although the Drosophila homologue of U2AF35 is required for viability18, there may be factors that are partially redundant with U2AF35. Urp is such a candidate because it is related to U2AF35 in structure and sequence. To test this possibility, we expressed and purified U2AF35 in the same way as Urp (Fig. 2b), and found that U2AF35 could not complement the Urp-depleted nuclear extract (Fig. 2c, lanes 7 to 9), indicating that Urp and U2AF35 are not functionally redundant. Results were identical when the complementation was repeated with a U2AF35 protein preparation that had previously been shown to be functional in restoring splicing activity to a U2AF35-depleted extract (ref. 5, and data not shown). Surprisingly, splicing in the Urp-depleted nuclear extract could be stimulated reproducibly by U2AF65(Fig. 2c, lanes 10 to 12). It has been found that U2AF65 can stimulate splicing in a U2AF35-deficient nuclear extract, probably as a result of recruiting the remaining U2AF35 in the depleted extract5. U2AF65 may therefore also interact with Urp, recruiting the remaining Urp in the Urp-depleted extract.
To test this, we examined the abundance of various splicing factors in the Urp-depleted nuclear extract by western blotting. As shown in Fig. 3a, Urp was effectively depleted by the anti-Urp affinity column in comparison to Urp present in untreated as well as preimmune serum-treated nuclear extracts. A fraction of U2AF65 was co-depleted, consistent with an interaction between U2AF65 and Urp. In contrast, the level of U2AF35 was only slightly reduced, indicating that a large fraction of U2AF35 is not associated with Urp or complexed with U2AF65, in agreement with previous functional studies on U2AF35 (ref. 5). The amount of SR proteins detected by monoclonal antibody 104 or of snRNP-associated proteins detected by anti-Sm antibodies was not affected by Urp depletion. The co-immunodepletion of Urp and U2AF65 indicates that Urp forms a functional complex with U2AF65in vivo. This conclusion does not contradict the observation that U2AF65 purifies only as a heterodimer with U2AF35, because 2M guanidine-HCl was used in a purification step19: these extreme conditions were not used in Urp depletion, so a weaker association between Urp and U2AF65 was likely to be preserved.
We also tested for any association of Urp with other splicing factors by western blotting analysis of high-salt eluants from the control and anti-Urp affinity columns. As expected, U2AF65, but not SR proteins or Sm antigens, was specifically detected in the eluant of the anti-Urp affinity column (Fig. 3b). U2AF35 seemed to be stoichiometrically present in the Urp–U2AF65 complex, because similar amounts of U2AF65 and U2AF35 were detected by specific antibodies of similar titres. This can be explained by the specific association of Urp with the U2AF heterodimer. Alternatively, Urp may be independently associated with equal amounts of U2AF65 and U2AF35, but we consider this less likely as only U2AF65 binds strongly to Urp in vitro and in the yeast two-hybrid system (see below).
We next assayed the binding between Urp and U2AF to investigate the role of Urp in splicing. We expressed GST–U2AF35 and GST–U2AF65 in bacteria, and carried out a GST-pulldown assay using in vitro translated Urp, U2AF35 and U2AF65(Fig. 4a). As expected, in vitro translated U2AF65 bound strongly to GST–U2AF35 immobilized on glutathione–Sepharose beads but not to vacant beads, demonstrating the specificity of binding. Under the same conditions (0.4M KCl), both Urp and U2AF35 bound efficiently to immobilized U2AF65(Fig. 4a). The interaction between Urp and U2AF65 was not mediated by RNA because the extent of binding was similar in the presence of RNase A (data not shown).
Interaction between Urp and U2AF65 could be mediated by the RS domain present in both proteins as this domain is a known protein–protein-interaction interface3,4, or by one or both of the U2AF35 homologous domains (H1 and H2). The H2 domain in Urp is a strong candidate because the corresponding region in U2AF35 is known to interact with U2AF65(ref. 20). We therefore prepared and tested a series of Urp mutants (Fig. 4b). Several conclusions can be drawn from the data shown in Fig. 4c and quantified in Fig. 4d. First, wild-type Urp binds to U2AF65, but its binding is less efficient than that between U2AF65 and U2AF35; the Urp–U2AF65 interaction was prevented in the presence of 1 M KCl (data not shown). This explains why Urp is associated with U2AF65 only under immunodepletion conditions and not under the more stringent purification conditions19. Second, deletion of the RS domain from Urp has no effect on its binding to U2AF65, indicating that the RS domain is not involved in the interaction between the two proteins. Third, although deletion of the H1 domain has no effect, deletion of the H2 domain significantly, but not completely, diminished Urp binding to U2AF65. Thus, U2AF35 and Urp seem to interact with U2AF65 through homologous sequences. However, the interaction between Urp and U2AF65 could not be competed by increasing amounts of unlabelled U2AF35 (data not shown), supporting the idea that Urp interacts with the U2AF heterodimer in vivo.
To verify the sequence requirements for protein–protein interaction between Urp and U2AF65, and to determine whether Urp, like U2AF35, also interacts with SR proteins through its RS domain, we extended the binding studies to the yeast two-hybrid system. As shown in Fig. 4e, Urp interacts with U2AF65 in yeast and deletion of the H2 domain diminishes binding, but deletion of the RS domain in Urp has no effect, consistent with the in vitro results. Urp, like U2AF35, interacts with two SR proteins, SC35 and SF2/ASF, but not with the U1-70k protein (data not shown); deletion of the RS domain in Urp reduces its interaction with either SR protein, indicating that the interaction between Urp and SR proteins is mediated by their RS domains (Fig. 4e). Deletion of the H2 domain had no effect on the interaction between Urp and SC35. However, for unknown reasons, the removal of the H2 domains from Urp decreased its interaction with SF2/ASF. We could not detect any significant interaction between U2AF35 and SF2/ASF, in contrast to a previous report3. Under the same conditions, Urp interacts with SF2/ASF, indicating a preference for SF2/ASF to bind with Urp rather than U2AF35(Fig. 4e).
We have shown that Urp is an essential splicing factor. On the basis of our finding that Urp is associated with the U2AF heterodimer in nuclear extracts and interacts with U2AF65 and SR proteins through distinct domains, we propose that Urp may form part of a larger U2AF complex and be engaged in network interactions during spliceosome assembly. Because Urp seems to play a non-redundant function with U2AF35 and appears preferentially to affect the second step of the splicing reaction, it may interact with a selective set of factors during reaction. The unique behaviour of Urp homologues in mouse indicates that Urp and its family members may be targets for regulation.
Cloning, expression and purification. SRPK1 was cloned in pGBT9 as bait to screen a HeLa cDNA library (from G. Hannon) constructed in the pGAD-GH vector as described21. The initial Urp clone was lacking 92 amino acids from its N terminus, and full-length Urp cDNA was isolated from a λgt11 HeLa S3 cDNA library (Clontech).
Urp was expressed as GST fusion proteins either in bacteria or by baculovirus. Thrombin-cleaved bacterial Urp was used to raise rabbit antibodies, diluted 1 : 3,000 for western blotting. For splicing complementation, baculovirus-expressed Urp and U2AF were purified from the inclusion bodies of Sf9 cells, followed by isolation from SDS–PAGE. Briefly, 108 infected Sf9 cells were sonicated in 10 ml buffer A (20 mM Tris, pH 7.6, 150 mM KCl, 2 mM EDTA, 1 mM DTT) plus standard protease inhibitors. Samples were centrifuged through a 15-ml sucrose cushion (40% sucrose in buffer A) at 12,000 r.p.m. for 30 min in a HB-4 rotor. Pellets were resuspended in 1 ml buffer A plus 8 M urea. An estimated 10 µg of specific protein was loaded per well onto a 10% SDS–polyacrylamide gel. Proteins were eluted from gel slices in buffer B (50 mM Tris, pH 7.5, 0.1% SDS, 0.1 mg ml−1 BSA, 1 mM DTT, 0.2 mM EDTA, 0.1 mM PMSF and 2.5% glycerol) overnight at 4 °C. Eluted proteins were precipitated with four volumes of cold acetone overnight at −20 °C. After a 15-min spin, the pellet was washed with cold methanol, dried and resuspended in 2.5 µl 8 M urea. Protein was renatured by adding 125 µl buffer C (20 mM Tris, pH 7.5, 10 mM KCl, 1 mM DTT, 20 mM PMSF and 0.2 mM EDTA), incubating at 4 °C for 24 h, and dialysing for 4 h against 20 mM Tris, pH 7.9, plus 1 mM DTT. Samples were normalized to contain 0.1 mg ml−1 protein, aliquoted, and stored at −80 °C.
Immunodepletion and reconstitution. Protein A–Sepharose CL-4B beads (0.5 ml; Pharmacia) were washed with PBS and mixed with an equal volume of anti-Urp serum or preimmune serum. After binding overnight, antibodies were crosslinked to protein A according to ref. 22. Antibody-bound beads were washed with 10 ml 0.1 M glycine (pH 2.5) then 2 × 10 ml buffer D (20 mM HEPES pH 7.6, 100 mM KCl, 1 mM DTT, 10% glycerol). HeLa cell nuclear extracts were used for immunodepletion as previously described5. To minimize co-depletion of other essential splicing factors, high-salt conditions were used: 1 ml nuclear extract was adjusted to 0.5 M KCl, then passed through a 0.5-ml (packed volume) antibody-affinity column 5 times at 4 °C. The depletion procedure was repeated once on a fresh 0.5-ml affinity column. Depleted extracts were dialysed against buffer D for 4 h at 4 °C, aliquoted, and stored at −80 °C. Antibody-affinity columns were regenerated by washing with 0.1 M glycine (pH 2.5) and stored in buffer D. Standard splicing reactions were carried out as described5.
In vitro and in vivo binding. U2AF65 and U2AF35cDNAs in pSP64 and in pGEX were provided by J. Fleckner and M. Green; anti-U2AF65 and U2AF35 antisera were gifts from P. Zuo and T. Maniatis, and were used at 1 : 5,000 dilution. Wild-type and mutant Urp cDNAs generated by PCR were inserted downstream of the T7 promoter in pcDNA3 (Invitrogen). In vitro-translated proteins were prepared using a TNT kit (Promega) in the presence of [35 S]methionine. 35S-labelled proteins were incubated with 1 µg GST–U2AF65 or GST–U2AF35 immobilized on 5 µl glutathione beads (Pharmacia) in binding buffer (20 mM HEPES, pH 7.6, 3 mM MgCl2, 0.1 mM EDTA, 0.1% Tween, 10% glycerol, 1 mM DTT, 0.4 M KCl) plus 5% BSA at 4 °C for 1 h. After beads were washed 4 times in binding buffer, proteins were resolved on SDS–PAGE, visualized by autoradiography, and quantified on a Phosphoimager.
Two-hybrid pairwise interactions were carried out using the LexA system23 in the host yeast strain EGY48. Bait plasmids contained individual cDNAs in frame downstream of the LexA gene in pEG202, and prey plasmids had individual cDNAs in pJG4-5. Several bait and prey plasmids (U2AF35 and SF2/ASF) were gifts from J. Wu3. EGY48 cells were co-transformed by the lithium acetate method and transformants were selected on glucose/CM-ura-his-trp plates. Six randomly picked colonies were restreaked on Gal/Raff/CM-ura-his-trp plates containing X-gal for colour development. Interactions were quantified by liquid β-galactosidase assay23.
Ruskin, B., Zamore, P. D. & Green, M. R. Afactor, U2AF, is required for U2 snRNP binding and splicing complex assembly. Cell 52, 207–219 (1988).
Krämer, A. The structure and function of proteins involved in mammalian pre-mRNA splicing. Annu. Rev. Biochem. 65, 367–490 (1996).
Wu, J. & Maniatis, T. Specific interactions between proteins implicated in splice site selection and regulated alternative splicing. Cell 75, 1061–1070 (1993).
Kohtz, J. D. et al. Protein–protein interactions and 5′-splice-site recognition in mammalian mRNA precursors. Nature 368, 119–124 (1994).
Zuo, P. & Maniatis, T. The splicing factor U2AF35 mediates critical protein-protein interactions in constitutive and enhancer-dependent splicing. Genes Dev. 10, 1356–1368 (1996).
Roscigno, R. F. & Garcia-Blanco, M. A. SR proteins escort the U4/U6.U5 tri-snRNP to the spliceosome. RNA 1, 692–706 (1995).
Fetzer, S., Lauber, J., Will, C. L. & Lührmann, R. The [U4/6.U5] tri-snRNP-specific 27K protein is a novel SR protein which can be phosphorylated by the snRNP-associated protein kinase. RNA 3, 344–355 (1997).
Gui, J. -F., Lane, W. S. & Fu, X. -D. Aserine kinase regulates intracellular localization of splicing factors in the cell cycle. Nature 369, 678–682 (1994).
Gui, J. -F., Tronchère, H., Chandler, S. D. & Fu, X. D. Purification and characterization of a kinase specific for the serine- and arginine-rich pre-mRNA splicing factors. Proc. Natl Acad. Sci. USA 91, 10824–10828 (1994).
Colwill, K. et al. SRPK1 and Clk/Sty protein kinases show distinct substrate specificities for serine/arginine-rich splicing factors. J. Biol. Chem. 271, 24569–24575 (1996).
Fu, X. -D. The superfamily of arginine/serine-rich splicing factors. RNA 1, 663–680 (1995).
Manley, J. L. & Tacke, R. SR proteins and splicing control. Genes Dev. 10, 1569–1579 (1996).
Kitagawa, K. et al. Isolation and mapping of human homologues of an imprinted mouse gene U2af1-rs1. Genomics 30, 257–263 (1995).
Nabetani, A., Hatada, I., Morisaki, H., Oshimura, M. & Mukai, T. Mouse U2af1-rs1 is a neomorphic imprinted gene. Mol. Cell. Biol. 17, 789–798 (1997).
Hyashizaki, Y. et al. Identification of an imprinted U2af binding protein related sequence on mouse chromosome 11 using the RLGS method. Nature Genet. 6, 33–40 (1994).
Yamaoka, T., Hatada, I., Kitagawa, K., Wang, X. & Mukai, T. Cloning and mapping of the U2af1-rs2 gene with a high transmission distortion in interspecific backcross progeny. Genomics 27, 337–340 (1995).
Zamore, P. D. & Green, M. R. Biochemical characterization of U2 snRNP auxiliary factor: an essential pre-mRNA splicing factor with novel intranuclear distribution. EMBO J. 10, 207–214 (1991).
Rudner, D. Z., Kanaar, R., Breger, K. S. & Rio, D. C. Mutations in the small subunit of the Drosphila U2AF splicing factor cause lethality and developmental defects. Proc. Natl Acad. Sci. USA 93, 10333–10337 (1996).
Zamore, P. D. & Green, M. R. Identification, purification, and biochemical characterization of U2 small nuclear ribonucleoprotein auxiliary factor. Proc. Natl Acad. Sci. USA 86, 9243–9247 (1989).
Zhang, M., Zamore, P. D., Carmo-Fonseca, M., Lamond, A. I. & Green, M. R. Cloning and intracellular localization of the U2 small nuclear ribonucleoprotein auxiliary factor small subunit. Proc. Natl Acad. Sci. USA 89, 8769–8773 (1992).
Hannon, G. J., Dmetrick, D. & Beach, D. Isolation of the Rb-related p130 through its interaction with CDK2 and cyclins. Genes Dev. 7, 2378–2391 (1993).
Harlow, E. & Lane, D. Antibodies: A Laboratory Manual(Cold Spring Harbor Laboratory Press, New York, (1988).
Zervos, A. S., Gyuris, J. & Brent, R. Mxi, a protein that specifically interacts with Max to bind Myc–Max recognition sites. Cell 72, 223–232 (1993).
We thank G. Hannon for yeast two-hybrid libraries; J.-L. Guan for raising anti-Urp antibodies; J. Fleckner and M. Green for U2AF cDNAs; P. Zuo and T. Maniatis for anti-U2AF antibodies and purified U2AF35 protein; J. Wu for two-hybrid plasmids and for advice; and members of our laboratory, especially H.-Y. Wang and J. Dyck, for suggestions and contributions. X.-D.F. is a Searle Scholar. H.T. was supported by postdoctoral fellowships from the Human Frontier Science Program and La Societé de Secours des Amis de la Science. J.W. was supported by a postdoctoral fellowship from the NIH. This work was supported by grants to X.-D.F. from the NIH.
About this article
Cite this article
Tronchre, H., Wang, J. & Fu, X. A protein related to splicing factor U2AF35 that interacts with U2AF65 and SR proteins in splicing of pre-mRNA. Nature 388, 397–400 (1997). https://doi.org/10.1038/41137
Minor Splicing Factors <i>Zrsr1</i> and <i>Zrsr2</i> Essential for Gametogenesis, Early Embryo Development and Conversion of Stem Cells into 2C-Like
SSRN Electronic Journal (2019)
Stem Cell Research & Therapy (2019)
Journal of Cellular and Molecular Medicine (2019)
From clonal hematopoiesis to myeloid leukemia and what happens in between: Will improved understanding lead to new therapeutic and preventive opportunities?
Blood Reviews (2019)
Dek42 encodes an RNA‐binding protein that affects alternative pre‐mRNA splicing and maize kernel development
Journal of Integrative Plant Biology (2019)