Protein composition of human prespliceosomes isolated by a tobramycin affinity-selection method

Plasmids, Oligonucleotides, and RNA Preparation.

Oligonucleotides used in this study were M13f (GTAAAACGACGGCCAGT), HPV23 (GATCCGGCTTAGTATAGCGAGGTTTAGCTACACTCGTGCTGAGCCGGATCCGCATG), and HPV24 (CGGATCCGGCTCAGCACGAGTGTAGCTAAACCTCGCTATACTAAGCCG). The J6f1 RNA aptamer (8) noncoding and coding sequences, respectively, are underlined. The EcoRI-BamH1 insert of pMINX (9) was first cloned into the pGEM-3Zf(+), yielding pGEM-MINX. The HPV23 and HPV24 oligonucleotides were annealed and cloned into BamH1/SphI linearized pGEM-MINX to yield pGEM MINX-T5. PCR with the HPV24 and M13f primers generated the transcription template for the aptamer-containing pre-mRNA. To obtain control pre-mRNA lacking the aptamer, the PCR product was cleaved with BamHI. m7G-capped transcripts were synthesized with T7 RNA polymerase (MegaScript, Ambion, Austin, TX). For quantitation, [α-³²P]UTP (3000 Ci/mmol; 1 Ci = 37 GBq) was added to 0.23 μM.

Solid-Phase Splicing Reaction and Spliceosome Purification.

N-hydroxysuccinimide-activated Sepharose 4 Fast Flow was derivatized with 5 mM tobramycin as described (10). All further procedures were performed at 4°C, unless otherwise stated. For affinity selection, 4× binding buffer (4× BP)/80 mM Tris⋅HCl, pH 9.1 at 4°C/4 mM CaCl₂/4 mM MgCl₂/2 mM DTT was freshly prepared. The tobramycin matrix was blocked with 400 μl of blocking buffer (1× BP/300 mM KCl/0.1 mg/ml tRNA/0.5 mg/ml BSA/0.01% Nonidet P-40) per 15-μl aliquot by head over tail (HOT) rotation overnight. The matrix was collected, and 400 μl of a mixture containing ≈60 pmol of tobramycin-tagged pre-mRNA and 40 μg of tRNA in binding buffer (1× BP/145 mM KCl), were added to 15 μl of matrix and incubated for 1.5 h. The matrix was then washed three times (750 μl each) with binding buffer containing 0.1% Nonidet P-40. A pre-mRNA concentration of 22–25 nM was found to be optimal for A complex formation (not shown). For a standard splicing assay, four 15-μl aliquots of matrix bound with pre-mRNA were prepared. Splicing mix (1.5 ml) was added per 15-μl matrix and incubated at the indicated time and temperature with constant HOT agitation. The splicing mix contained 35% HeLa cell nuclear extract (11) in buffer D [20 mM Hepes-KOH, pH 7.9/100 mM KCl/1.5 mM MgCl₂/0.2 mM EDTA, pH 8.0/0.5 mM DTT/0.5 mM PMSF/10% (vol/vol) glycerol] and was supplemented with 25 mM KCl/3 mM MgCl₂/2 mM ATP/20 mM creatine phosphate. After splicing, reactions were chilled on ice, and the matrix was collected by centrifugation. Matrix aliquots were pooled and washed three times with 750 μl of washing buffer (1× BP/75 mM KCl/0.1% Nonidet P-40). Complexes were eluted by incubating with 250 μl of elution buffer (1× BP/5 mM tobramycin/145 mM KCl/2 mM MgCl₂) with HOT rotation for 10 min. Typically, ≈75% of the bound complexes could be eluted. RNA and protein were recovered and analyzed on 8.3 M urea/10% polyacrylamide gels (RNA) or by SDS/10–13% PAGE (proteins), or for splicing, on 8.3 M urea-15% polyacrylamide gels.

Glycerol Gradient Centrifugation.

Tobramycin eluate (220 μl) was loaded onto a 3.8-ml linear 10–30% glycerol gradient in 20 mM Hepes-KOH, pH 7.9/75 mM KCl/1.5 mM MgCl₂/0.5 mM DTT. Gradients were centrifuged for 107 min at 60,000 rpm in a Sorvall TH660 rotor and harvested manually in 175-μl fractions from the top. RNA and protein were extracted from 100 μl of every odd fraction and analyzed as above.

Psoralen Crosslinking and Northern Blotting.

Crosslinking was initiated by adding psoralen (4′-aminomethyl-4,5′,8-trimethylpsoralen hydrochloride) in DMSO (2 mg/ml) to 75 μl of the even gradient fractions to a final concentration of 40 μg/ml and incubating for 10 min on ice. Samples were irradiated with 365 nm light for 30 min at 4°C. RNA was recovered (12) and resolved on a 5% polyacrylamide/8.3 M urea gel and analyzed by Northern blotting essentially as described (13). DNA probes were generated by random priming (Prime-it II, Promega) using pMINX DNA (pre-mRNA) or PCR-amplified U1 or U2 snRNA coding sequences and [α-³²P]dATP (6,000 Ci/mmol).

MS.

Proteins separated by SDS/PAGE were analyzed by matrix-assisted laser desorption/ionization MS (MALDI-MS) or liquid chromatography-coupled tandem MS [LC-MSMS as described in ref. 14; for details, see Supporting Methods, which is published as supporting information on the PNAS web site, www.pnas.org) and identified in the National Center for Biotechnology Information nonredundant database by using MASCOT (Matrix Science, London) as a search engine.

Tobramycin Affinity Selection of Native Spliceosomal Complexes.

To affinity-purify spliceosomes under native conditions, we introduced a 40 nt RNA aptamer at the 3′ end of the adenovirus-derived MINX pre-mRNA. This aptamer binds with high affinity (5 nM) the aminoglycoside antibiotic tobramycin under physiological conditions (8). Indeed, the aptamer-tagged MINX pre-mRNA bound efficiently to tobramycin-derivatized Sepharose beads when the binding reaction was performed at pH 9.1 in buffer containing 145 mM KCl and 1 mM Mg²⁺. Under these conditions, 60–70% of aptamer-tagged pre-mRNA, but only ≈2% of untagged pre-mRNA, bound. More importantly, 80% of the bound pre-mRNA could be eluted by incubating with 5 mM tobramycin. The aptamer-tagged MINX pre-mRNA formed spliceosomal complexes and was spliced in vitro with efficiencies and kinetics similar to that of untagged MINX pre-mRNA (not shown). However, in the presence of nuclear extract, <1% of the tagged pre-mRNA bound to the tobramycin matrix, suggesting that the RNA aptamer is no longer accessible or the conditions used for splicing are suboptimal for binding. Solid-phase splicing was thus performed by first immobilizing the tagged pre-mRNA on the tobramycin matrix and subsequently incubating with nuclear extract under splicing conditions (Fig. 1A). The matrix was then washed extensively, and bound material was eluted with tobramycin. RNA was isolated from the eluate or reaction supernatant and analyzed by denaturing PAGE (Fig. 1B). Splicing intermediates/products were first observed in those complexes eluted from the matrix after 90 min and increased markedly after 180 and 270 min, when mRNA product became readily detectable (Fig. 1B, lanes 2–4). Thus, splicing proceeds with significantly reduced kinetics as compared with the nonimmobilized, tagged MINX pre-mRNA, which yielded splicing products already after 15 min (not shown). Most of the pre-mRNA remained bound to the matrix even after extended incubation times, as evidenced by the presence of only a very low level (maximally 7%) of pre-mRNA and splicing intermediates in the supernatant of the solid-phase splicing reaction (Fig. 1B, lanes 5–7). However, consistent with it's release from spliceosomes after the second step of splicing, elevated amounts of excised intron were detected. These results demonstrate that despite reduced reaction kinetics, the matrix-bound, aptamer-tagged pre-mRNA is spliced efficiently and, therefore, functional spliceosomal E, A, B, and C complexes must be assembled as well.

Purification of Spliceosomal Complex A.

To isolate prespliceosomes, we performed solid-phase splicing for only 45 min to minimize formation of B/C complexes. The RNA and protein content of the tobramycin eluate was then analyzed by denaturing PAGE or SDS/PAGE, respectively. As a control for nonspecific interactions, an affinity selection was also performed in an identical manner, but without pre-mRNA. As shown in Fig. 2A, in the absence of pre-mRNA, relatively little snRNA was detected in the eluate (lane 2). By contrast, the eluate of the aptamer-tagged pre-mRNA (lane 1) contained similar amounts of pre-mRNA, U1 and U2 snRNA, and only a low level of U4, U5, and U6 snRNA. This result indicated that predominantly splicing complex A (and/or potentially E) had formed after 45 min and was subsequently eluted with tobramycin. Correspondingly, the eluate from the affinity selection performed with pre-mRNA exhibited a complex but distinct protein pattern, whereas only a few proteins migrating in the 55- to 65-kDa range were observed without pre-mRNA (Fig. 2A, compare lanes 3 and 4).

To characterize the affinity-purified splicing complexes in more detail, the eluate was subjected to glycerol gradient centrifugation (Fig. 2B), and the sedimentation behavior of complexes containing ³²P-labeled pre-mRNA was compared with that of naked pre-mRNA or affinity-selected H complexes (i.e., complexes formed after a 15-min incubation with nuclear extract at 4°C). The vast majority of complexes in the 45-min eluate peaked in fraction 11, with a sedimentation value of ≈22S; minor peaks sedimenting at ≈28–42S were also observed. In contrast, naked pre-mRNA and H complexes sedimented at ≈12 and ≈13S, respectively. Thus, the 45-min eluate seems to consist of a relatively homogeneous population of splicing complexes which exhibit an S value significantly higher than the H complex, which is consistent with the idea that predominantly complex A or E had been isolated. Indeed, analysis of the RNA composition of the gradient fractions from the 45-min eluate revealed equimolar amounts of the pre-mRNA and U1 and U2 snRNAs in fractions 9–13 (Fig. 2C). These fractions were nearly free of U4, U5, and U6 snRNAs, which, for the most part, were found in either slower or faster migrating complexes; the latter, which contain equimolar amounts of the pre-mRNA and all of the major spliceosomal snRNAs (lanes 17–23), likely correspond to spliceosomal complex B. Thus, glycerol gradient centrifugation allows the efficient separation of H and B complexes from the affinity-selected A/E complexes.

Identification of E and A Complexes via RNA–RNA Crosslinking.

E and A complexes both contain U1 and U2 snRNPs, but a U2/pre-mRNA base pairing interaction is first observed in the A complex (3). Thus, RNA–RNA crosslinking (12) was performed to identify precisely complexes migrating in the 22S region of the gradient. Gradient fractions were UV-irradiated in the presence of psoralen, and the RNA was analyzed by Northern blotting, sequentially probing for the pre-mRNA, U1 and U2 snRNAs (Fig. 3 A–C). Multiple U1 snRNA-containing crosslinks, which peaked in fraction 10, were observed in fractions 8–14 (Fig. 3C). Each of these bands was also detected with a probe specific for the pre-mRNA, confirming that they represent crosslinks between the splicing substrate and U1 (Fig. 3A). In contrast, a U2/pre-mRNA crosslink was first detected in fraction 12, and its intensity gradually diminished from fraction 14 to 22 (Fig. 3B). In an independent experiment, U2/pre-mRNA crosslinks were also observed in fraction 11, but not in the preceding fractions (not shown). Thus, fractions 11–14 contain predominantly the spliceosomal A complex, whereas fractions 8–10 likely contain mainly E complex. A major U2/U6 crosslink (verified by probing for U6; not shown), which is indicative of complex B, peaked in fractions 16–22. However, only small amounts were detected below fraction 14, confirming that fractions 11–13 (the A complex peak) are essentially free of complex B.

Protein Composition of Prespliceosomes.

The protein composition of the various gradient fractions was analyzed by SDS/PAGE (Fig. 4A). A differential distribution of proteins across the gradient was observed, with high molecular weight U2 and U5 proteins peaking in fractions 9–13 and 17–23, respectively. In addition, a distinct protein pattern was found in the uppermost fractions, which essentially lack pre-mRNA, suggesting that some proteins (or snRNPs; see also Fig. 2C for RNA composition) have dissociated during elution and/or gradient centrifugation. Proteins present in fractions 11–13 (the A complex peak) were pooled, fractionated by SDS/PAGE, and analyzed by MALDI-MS or, additionally, by LC-MSMS.

More than 70 proteins were identified, including almost all proteins known to be present in 12S U1 and 17S U2 snRNPs (Fig. 4B and Table 1). In contrast, no U4/U6 and only high molecular weight U5 and/or tri-snRNP proteins were detected, confirming the absence of large amounts of 25S U4/U6.U5 tri-snRNPs. Consistent with their known roles during E/A complex formation, most members of the SR protein family and the essential factors U2AF65 and U2AF35 were identified. A large number of hnRNP proteins and both m7G-cap-binding proteins (CBP20 and CBP80) were also identified. In addition to the 17S U2 snRNP-associated proteins hPrp5 and hPrp43, two more members of the DExH/D-box family of RNA unwindases/RNPases, namely, RNA helicase A (DDX9) and p68, were detected. Several proteins implicated in pre-mRNA splicing that appear to perform multiple roles in the cell, such as Y-box-binding protein 1 (YB-1), TLS/FUS, and NFAR-2 (15–19), were identified, as well as the mRNA export factors Aly and HuR (20–22). Interestingly, affinity-purified prespliceosomes also contained a number of proteins not previously known to be involved in splicing (Table 1).