Intron Removal Requires Proofreading of U2AF/3' Splice Site Recognition by DEK

See allHide authors and affiliations

Science  30 Jun 2006:
Vol. 312, Issue 5782, pp. 1961-1965
DOI: 10.1126/science.1128659


Discrimination between splice sites and similar, nonsplice sequences is essential for correct intron removal and messenger RNA formation in eukaryotes. The 65- and 35-kD subunits of the splicing factor U2AF, U2AF65 and U2AF35, recognize, respectively, the pyrimidine-rich tract and the conserved terminal AG present at metazoan 3′ splice sites. We report that DEK, a chromatin- and RNA-associated protein mutated or overexpressed in certain cancers, enforces 3′ splice site discrimination by U2AF. DEK phosphorylated at serines 19 and 32 associates with U2AF35, facilitates the U2AF35-AG interaction and prevents binding of U2AF65 to pyrimidine tracts not followed by AG. DEK and its phosphorylation are required for intron removal, but not for splicing complex assembly, which indicates that proofreading of early 3′ splice site recognition influences catalytic activation of the spliceosome.

Aminimal U2AF heterodimer consisting of RNA recognition motifs (RRM) 1 and 2 of U2AF65 (1) and the U2AF homology motif (UHM or ΨRRM) of U2AF35 (2) was analyzed by nuclear magnetic resonance (NMR) spectroscopy in the absence or presence of an RNA containing a pyrimidine tract followed by a consensus 3′ splice site (3′ ss) [5′ (U)13ACAGG 3′]. As expected from the affinity of U2AF65 for uridine-rich sequences (1), the presence of the RNA caused extensive changes in the NMR spectrum of the U2AF65 RRM 1+2 subunit (Fig. 1A, left). In contrast, small perturbations concerning few residues were observed in the U2AF35 ΨRRM spectrum (right). The latter was unexpected, because previous observations suggested that U2AF35 specifically recognizes the 3′ss-AG (35). Gel retardation assays using 32P-uridine–labeled RNAs [5′ GGG(U)13AC-AG/CG-GUAAAAUAACUCA 3′] showed that, although U2AF35 ΨRRM increases the affinity of the complex threefold, the effect is similar for AG-, CG-, UG- or AA-3′ss, strong or weaker pyrimidine tracts (Fig. 1B and figs. S1 and S2). Lack of AG discrimination was also observed when different assays and recombinant full-length U2AF heterodimer or U2AF purified from HeLa cells were utilized (Figs. 1, C and D). In contrast, both endogenous U2AF and the minimal heterodimer showed preferential ultraviolet (UV) light–induced photo–cross-linking of U2AF65 to AG-3′ss RNAs in nuclear extracts (Fig. 1E). Reconstitution of U2AF-depleted extracts with recombinant U2AF subunits indicated that U2AF35 is required for AG discrimination (Fig. 1E, bottom). The presence of U2AF35 and other components of the nuclear extract decreased cross-linking of U2AF65 to the nonconsensus CG-3′ss, which suggests the existence of a proofreading activity that enforces specific association of U2AF with pyrimidine tracts followed by consensus AG-3′ss.

Fig. 1.

An activity present in nuclear extracts is necessary for 3′ss AG discrimination by U2AF. (A) Two-dimensional NMR spectra of 1H, 15N amide resonance correlations of U2AF65 RRM 1+2 (left) and U2AF35 ΨRRM (right) forming a minimal U2AF heterodimer. Spectra in the absence (black) or presence (red) of a twofold molar excess of 5′(U)13ACAGG 3′ RNA are shown. (B) Gel retardation assays using the indicated proteins (10–6 M for lanes 1 and 8, and serial threefold dilutions) and 10 fmol of either 5′ 32P-labeled RNA used in (A) (AG) or a mutant replacing the 3′ss AG by CG (CG). Positions of unbound RNA and complexes are indicated. (C) UV-induced cross-linking assay using GST-U2AF65 (2 × 10–7 M), His-U2AF35 (5 × 10–7 M) and 32P-uridine–labeled 5′ GGG(U)13AC-CG/CG-GUAAAAUAACUCA 3′ RNAs (10–9 M). (D) Gel retardation assay using purified U2AF heterodimer from HeLa cells (10–8 M for lanes 2 and 7, and serial twofold dilutions). (E) UV-induced cross-linking and immunoprecipitation of endogenous U2AF and truncated heterodimer subunits (2.5 × 10–5 M) in the presence of HeLa (NE) or U2AF-depleted (ΔU2AF NE) nuclear extracts (33% of reaction volume). Ratios between cross-linking intensities for AG- and CG-containing RNAs are indicated (n = 3).

This activity cofractionated with U2AF during the two first chromatographic steps of U2AF purification (6) (fig. S3). In fig. 2A, compare lanes 3 and 4 with 7 and 8 for the U2AF-containing complex (identified in lane 2 by supershift with antibodies against U2AF65). The activity was, however, separated from U2AF on the next chromatographic step [poly(U)-Sepharose]; whereas U2AF was retained in the column (6), the flow-through fraction provided AG versus CG discrimination to the truncated heterodimer in both UV-mediated cross-linking (Fig. 2B) and gel-retardation assays (fig. S4). The activity present in this fraction was retained on an affinity column containing the truncated U2AF heterodimer (Fig. 2D, lanes 1 to 4). Comparison of the protein profiles of the input and flow-through fractions revealed that a 50-kD protein was retained in the U2AF column (Fig. 2C, lower component of the 50-kD doublet). Mass spectrometry analyses identified this protein as DEK, a chromatin-, pre–mRNA- and mRNA-associated protein overexpressed or mutated in certain cancers (7, 8). Consistent with a role for DEK in providing AG discrimination to U2AF, depletion of DEK from HeLa nuclear extracts (fig. S5) resulted in reduced AG versus CG discrimination by endogenous U2AF65 (Fig. 2E, lanes 1 to 4), an effect that was reversed when recombinant purified DEK was added to the depleted extracts (lane 5). Cross-linking between U2AF35 and an RNA radioactively labeled at the 3′ss dinucleotide (A-[32P]-G or C-[32P]-G) was reduced in DEK-depleted extracts, which indicated that DEK is required for 3′ss recognition by U2AF35 (Fig. 2F). Collectively, the results described above indicate that DEK provides a proofreading function that allows U2AF to discriminate between bona fide AG-containing and nonconsensus 3′ss regions.

Fig. 2.

DEK enforces discrimination by U2AF between 3′ss AG and CG. (A) Gel retardation assay using the DEAE Sepharose, phosphocellulose (PC) eluate of U2AF purification procedure (6) carried out as in Fig. 1B. The position of the U2AF-containing complex (as shown by supershift with antibody against U2AF65) is indicated. (B) UV-induced cross-linking and immunoprecipitation of U2AF65 RRM 1+2 in the presence of the indicated chromatographic fractions, carried out as in Fig. 1E. pUS FT, flow-through of poly(U) Sepharose loaded with the fraction tested in (A). pUS 250 mM KCl, eluate of the same column at 250 mM KCl. (C) Silver stain of protein profiles of chromatographic fractions described in (B), before and after U2AF affinity chromatography. The asterisk indicates proteins leaked from the affinity column. (D) UV-induced cross-linking and immunoprecipitation of U2AF65 RRM 1+2 in the presence of the indicated chromatographic fractions and recombinant proteins (2.5 × 10–6 M). Input: pUS FT, as in (B); FT, material unbound to U2AF affinity column; wt, GST-DEK; 19,32 A and 19,32 D: DEK mutants with serine replaced by alanine (A) or aspartic acid (D) at these positions. ATP was present in all assays except where indicated. Ratios between cross-linking intensities for AG- and CG-containing RNAs are indicated (n = 3). (E) UV-induced cross-linking and immunoprecipitation assays as in (B), using mock- or DEK-depleted nuclear extracts, complemented with 100 ng/μl of purified recombinant DEK or mutant derivatives. (F) Cross-linking and immunoprecipitation assays of U2AF35 carried out as in (E) using RNAs 32P-labeled at ApG or CpG 3′ss and antibodies against U2AF35.

DEK retention in U2AF affinity columns suggested the possibility of an interaction between these factors. Pull-down experiments using in vitro translated, 35S-labeled U2AF65 or U2AF35 and recombinant purified glutathione S-transferase (GST)-DEK revealed formation of a complex between DEK and U2AF35, which was, at least in part, RNA-independent and involved the 100 amino-terminal residues of DEK (Fig. 3, A and B). Interestingly, the interaction was disrupted by phosphatase treatment (Fig. 3B, lanes 3 versus 4 and 11 versus 12), which suggests the requirement for protein phosphorylation. Indeed, DEK is a phosphoprotein (9), and phosphorylation occurs upon incubation of DEK(1–100) with rabbit reticulocyte lysates (fig. S7). Mutation of serines 19 and 32 to alanine (A) or aspartic acid (D) abolished phosphorylation (fig. S6). Interaction with U2AF35 was abolished by mutation to alanine; mutation to the phosphorylation mimic, aspartic acid, allowed the interaction to occur even in the presence of phosphatase (Fig. 3B, lanes 5 to 8 and 13 to 16). Taken together, the results of Fig. 3 reveal formation of a complex involving the amino-terminal 100 amino acids of DEK and U2AF35, which is dependent on phosphorylation of serines 19 and 32. The involvement of additional factors in complex formation cannot be ruled out, because NMR experiments failed to detect significant DEK-induced changes in U2AF35 structure or interaction with RNA.

Fig. 3.

Phosphorylation-dependent association between DEK and U2AF35. (A) Pull-down of 35S-labeled in vitro translated U2AF subunits by the indicated GST fusion proteins (PUF-60 and TIA-1 are splicing factors used as negative controls). P, precipitation products; SN, 1/10 of nonprecipitated material. Ribonuclease (RNase) treatment and positions of in vitro translated products are also indicated. (B) Pull-down experiments as in (A), using GST fusions of DEK (residues 1 to 100) and the indicated mutant derivatives. CIP indicates treatment with calf intestine alkaline phosphatase.

Recombinant DEK was found insufficient to confer discrimination to the purified heterodimer (Fig. 2D, lanes 5 and 6). Discrimination was restored, however, when DEK was combined with the DEK-depleted flow-through fraction from the U2AF affinity column mentioned above (Fig. 2D, lanes 7 and 8), but not with other chromatographic fractions (fig. S3B). The requirement for adenosine triphosphate (ATP) (see below), for U2AF35 (Fig. 2D, lanes 9 and 10) and the phosphorylation-dependent association of U2AF35 suggested the possibility that DEK phosphorylation by a kinase present in the complementing fraction was necessary for AG discrimination by U2AF. Consistent with this hypothesis, DEK was phosphorylated in the presence of the complementing fraction (fig. S6), and mutation of serines 19 and 32 to alanine compromised the discriminatory activity provided by DEK, whereas mutation to aspartic acid maintained the activity in reconstituted assays (Fig. 2D, lanes 11 to 14) and in complementation of DEK-depleted extracts (Fig. 2E, lanes 6 and 7). Part of the discrimination provided by the mutant with aspartic acid at positions 19 and 32 was lost in the absence of ATP (lanes 15 and 16), which suggests that phosphorylation of serines 19 and 32 is necessary, but not sufficient, for full DEK activity. Collectively, the results of Figs. 2 and 3 indicate that DEK phosphorylation facilitates its association with U2AF35, and this correlates with the activity of the protein to confer AG discrimination to the U2AF heterodimer.

One possible consequence of the absence of DEK proofreading would be splicing activation of pre-mRNAs containing mutations at the 3′ss AG. This was found not to be the case for either IgM pre-mRNA (which requires the 3′ss AG to undergo both catalytic steps of splicing) or AdML (which undergoes—at least to some extent—the first catalytic step in the absence of 3′ss AG) (Fig. 4A). DEK depletion, however, inhibited the catalytic steps requiring AG recognition in wild-type RNAs: first step for the AG-dependent and, mainly, second step for the AG-independent pre-mRNAs (Fig. 4B). Splicing was restored by complementation with recombinant purified DEK and the DEK mutant with Asp19 and Asp32, but not with the DEK mutant with Ala19 and Ala32 (Fig. 4, B and C, and fig. S7). U2AF was not depleted in DEK-depleted extracts (fig. S5), and splicing was not restored by an excess of U2AF65 or U2AF heterodimer (Fig. 4C). Collectively, these data establish a correlation—further substantiated by the common requirement for phosphorylation of Ser19 and Ser32—between three properties of DEK: associating with U2AF35, proofreading U2AF/3′ ss recognition, and sustaining pre-mRNA splicing.

Fig. 4.

DEK is required for pre-mRNA splicing but not for spliceosome assembly. (A) In vitro splicing assays using mouse immunoglobulin IgM or adenovirus major late (AdML) promoter pre-mRNAs in which the 3′ss AG was mutated to GA, in mock- or DEK-depleted nuclear extracts. (B) Splicing assays as in (A) using wild-type pre-mRNAs, in the absence or presence of recombinant purified DEK (30 ng/μl). The positions of pre-mRNA, splicing intermediates, and products are indicated. AdML (bottom) shows a clearer effect on lariat intermediates and products. (C) Splicing complementation assays as in (B), recombinant DEK; DEK Ala (A) and Asp (D) mutants at serines 19 and 32; recombinant U2AF65 (5, 10, and 15 ng/μl); and purified HeLa U2AF heterodimer (10 ng/μl). Quantification of splicing efficiency from three independent experiments is shown. (D) Spliceosome assembly corresponding to splicing assays as in (B). IgM 3′, RNA containing 3′ half intron and exon 2 of IgM pre-mRNA. Positions of heterogeneous nuclear RNP (H) and spliceosomal (A and B/C) complexes are indicated. (E) Spliceosome assembly assays as in (D) using IgM 3′ RNA in the presence of 50, 200, and 400 molar excess of unlabeled RNAs containing AG or CG at the 3′ss as in Fig. 1B.

In contrast, neither U2 small nuclear ribonucleoprotein (snRNP) binding nor subsequent events in spliceosome assembly were affected by DEK depletion (Fig. 4D), which suggests that the absence of proofreading at early stages of 3′ss recognition influences catalytic activation of splicing complexes. Effects of DEK depletion on splicing complex formation were, however, observed in competition assays in which U2 snRNP binding to a 3′ss-containing RNA was competed by an excess of RNAs containing a pyrimidine tract followed by AG or CG. Although only AG-containing RNAs were able to compete in mock-depleted extracts, both AG- and CG-containing RNAs were effective competitors in DEK-depleted extracts (Fig. 4E and fig. S8). These results are consistent with the notion that DEK enforces selective functional association of U2AF with bona fide 3′ss.

Alterations in splice-site recognition and intron removal could underlie the molecular basis of pathologies involving DEK overexpression or inactivation, including autoimmune disease and cancer (7). Consistent with this, overexpression of mutant versions of DEK in cells in culture results in reduced splicing efficiency of model pre-mRNAs (fig. S9). Previous data from iterative selection of sequences from a random pool (3) indicated that the U2AF heterodimer is sufficient to select both pyrimidine tracts and consensus 3′ss AG. DEK may enhance this intrinsic preference, revealed to different extents depending on assay stringency. DEK binding could induce a conformational change in U2AF35 that enforces AG selectivity (Fig. 2F), similar to the enhanced specificity of U2A′ for U2 snRNA sequences upon interaction with U2B″ (10). Elegant studies in yeast implicated the transient interaction of the RNA-dependent adenosine triphosphatase (ATPase) Prp16 with the spliceosome as a timing device that enforces kinetic proofreading of lariat intermediates before the second catalytic step (11). An RNA-dependent ATPase could facilitate DEK-mediated displacement of U2AF from pyrimidine tracts not followed by AG 3′ss, as reported for other RNA-protein interactions (12). Kinetic proofreading could explain why catalytic activation of the spliceosome requires discrimination between bona fide and cryptic 3′ss during early splicing complex formation. Phosphorylation affects DEK activity in 3′ss recognition and splicing, which suggests that casein kinase II (13) or other kinases can influence the selectivity of 3′ss identification and possibly alternative splicing. Phosphorylation could also cause a switch in DEK partners, including those that mediate its association with chromatin or the transcriptional machinery (7); this may underlie aspects of the coupling between transcription and pre-mRNA processing (14).

Supporting Online Material

Materials and Methods

Figs. S1 to S9

References and Notes

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article