Portal de Periódicos da CAPES

Site-specific deoxynucleotide substitutions in yeast U6 snRNA block splicing of pre-mRNA invitro

1997; Springer Nature; Volume: 16; Issue: 8 Linguagem: Inglês

10.1093/emboj/16.8.2119

1460-2075

Chang Hee Kim,

RNA modifications and cancer

Article15 April 1997free access Site-specific deoxynucleotide substitutions in yeast U6 snRNA block splicing of pre-mRNA in vitro Chang Hee Kim Chang Hee Kim Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Daniel E. Ryan Daniel E. Ryan Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Tadeusz Marciniec Tadeusz Marciniec Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author John Abelson Corresponding Author John Abelson Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Chang Hee Kim Chang Hee Kim Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Daniel E. Ryan Daniel E. Ryan Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Tadeusz Marciniec Tadeusz Marciniec Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author John Abelson Corresponding Author John Abelson Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA Search for more papers by this author Author Information Chang Hee Kim1, Daniel E. Ryan1, Tadeusz Marciniec1,2 and John Abelson 1 1Division of Biology 147-75, California Institute of Technology, Pasadena, CA, 91125 USA 2The Victor Chang Cardiac Research Institute, St Vincent's Medical Centre, 376 Victoria Street, Darlinghurst, NSW 2010 Australia *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:2119-2129https://doi.org/10.1093/emboj/16.8.2119 Retraction(s) for this article Site-specific deoxynucleotide substitutions in yeast U6 snRNA block splicing of pre-mRNA in vitro17 May 1999 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info We have identified 2′-hydroxyl groups of the U6 phosphate-ribose backbone which are required for reconstitution of splicing activity in U6-depleted yeast extract. To screen the 2′-hydroxyls of yeast U6 at nucleotides 39–88, spanning the conserved central domain, synthetic U6 RNAs were constructed with deoxyribonucleotides incorporated site specifically. Only four individual deoxynucleotide substitutions blocked splicing activity: dA51 (in the ACAGAG sequence), dA62 (next to the AGC triad), and dU70 and dC72 (both in the loop of the 3′ intramolecular stem–loop). Native gel analysis revealed that these deoxy-substituted U6 RNAs were competent for assembly of spliceosomes. Interestingly, a 2′-O-methyl substituent at A51, A62, U70 or C72 did not inhibit splicing activity, indicating that the essential 2′-OH groups at these positions in U6 act as hydrogen bond acceptors or neutral coordinated ligands. The requisite 2′-hydroxyls at A62, U70 and C72 show both similarities and differences relative to the positions of essential 2′-hydroxyls of catalytic domain V of group II ribozymes. The identification of the essential 2′-hydroxyls at positions 62, 70 and 72 corroborates that the 3′ intramolecular stem–loop in U6 plays an important role in pre-mRNA splicing. Introduction U6 snRNA is an essential component of fully assembled spliceosomes—large (40S) ribonucleoprotein complexes that mediate the process of pre-mRNA splicing (for reviews, see Madhani and Guthrie, 1994b; Nilsen, 1994; Sharp, 1994; Ares and Weiser, 1995). The central domain of U6 (nucleotides 47–85 in yeast) is highly conserved from yeast to mammals (Brow and Guthrie, 1988; Roiha et al., 1989; Shumyatsky and Reddy, 1993). Two-thirds of the conserved U6 domain is base-paired with U4 snRNA in U4–U6 snRNP complexes (Bringmann et al., 1984; Hashimoto and Steitz, 1984; Rinke et al., 1985; Brow and Guthrie, 1988; Vankan et al., 1990); the U4–U6 snRNPs deliver U6 to a spliceosome during the ordered process of spliceosome assembly (Pikielny et al., 1986; Cheng and Abelson, 1987; Konarska and Sharp, 1987). Near completion of assembly, the U4–U6 duplex is dissociated, and U4 may leave the spliceosome (Pikielny et al., 1986; Cheng and Abelson, 1987; Lamond et al., 1988; Yean and Lin, 1991). Dissociation of U4–U6 allows a 5′ region of the U6 conserved domain to base-pair with U2 snRNA, thus forming U2–U6 helix I in fully assembled spliceosomes (Madhani and Guthrie, 1992) (see model of the fully assembled yeast spliceosome, Figure 1). Concurrently, the 3′ region of the U6 conserved domain putatively forms an intramolecular stem–loop as present in non-complexed U6 snRNPs in vivo (Wolff and Bindereif, 1991; Fortner et al., 1994; Brow and Vidaver, 1995). The conserved ACAGAG region of U6, nucleotides 47–52 in yeast, makes physical contact with the 5′ splice site of pre-mRNA, as revealed in cross-linking experiments (Sawa and Abelson, 1992; Sawa and Shimura, 1992; Wassarman and Steitz, 1992; Sontheimer and Steitz, 1993; Kim and Abelson, 1996). If U6 is associated simultaneously with U2 (at helix I) and pre-mRNA (at the 5′ splice site) while U2 is base paired to the pre-mRNA at the branch site (Parker et al., 1987), then the pre-mRNA could be folded to juxtapose the 5′ splice site and the nucleophilic bulged adenosine of the branch site, perhaps thereby activating the first catalytic step of splicing (cf. Figure 1). Interactions between a conserved stem–loop of U5 snRNA and both the 5′ and 3′ splice sites are thought to position the 5′ and 3′ exons for ligation at the second catalytic step of splicing (Newman and Norman, 1991, 1992; Wyatt et al., 1992; Sontheimer and Steitz, 1993; Frank et al., 1994). The snRNA components of spliceosomes are thought to function as catalysts for the two reactive steps of pre-mRNA splicing. This catalytic RNA hypothesis is founded on the knowledge that the group II self-splicing intron, an RNA-only system (reviewed by Michel and Ferat, 1995), has intronic sequences that catalyze self-splicing via the same phosphodiester transfer reactions as occur in pre-mRNA splicing, including the formation of a 2′-5′ phosphodiester-linked lariat intermediate. Figure 1.Model of the fully assembled yeast spliceosome. Interactions between U2, U6 and pre-mRNA are illustrated. Mutationally sensitive nucleotides, essential pro-R phosphate oxygen atoms, essential 2′-hydroxyl groups (this work) and U6 cross-linking interactions are highlighted. Download figure Download PowerPoint The structure–function relationships of U6 in the spliceosome are not well understood but are beginning to be elucidated. Specific mutations of individual nucleotides in U6 block the first catalytic step of splicing and prevent spliceosome assembly; other mutations block the second catalytic step (Fabrizio and Abelson, 1990, 1992; Vankan et al., 1990; Madhani and Guthrie, 1992). In fully assembled spliceosomes, U2–U6 helix I was identified by compensatory mutations and was demonstrated to be essential for splicing activity in yeast (Madhani and Guthrie, 1992). Two other U2–U6 helices, flanking the conserved U6 domain, were similarly identified and found to be important for splicing in mammalian cells (Hausner et al., 1990; Datta and Weiner, 1991; Wu and Manley, 1991; Wolff and Bindereif, 1992; Sun and Manley, 1995), but are not important in yeast (Fabrizio et al., 1989; Madhani et al., 1990; Yan and Ares, 1996). A randomization–selection experiment of both U2 and U6 snRNAs provided evidence of a tertiary contact between U2 and U6 in the helix I region using a method analogous to phylogenetic co-variation analysis (Madhani and Guthrie, 1994a). Genetic suppression studies of 5′ splice site mutants of pre-mRNA identified base-paired interactions between the 5′ splice site and U6 that are important for 5′ splice site recognition (Kandels-Lewis and Séraphin, 1993; Lesser and Guthrie, 1993). U6 structure–function relationships are key features of the fully assembled yeast spliceosome (Madhani and Guthrie, 1994b) (cf. Figure 1). In addition to the U6 sequence and secondary structural requirements for splicing activity, other essential structural features of U6 reside on the phosphate-ribose backbone. Important phosphate oxygen atoms of U6 were identified by incorporating thiophosphate groups at selected positions in the U6 sequence (Fabrizio and Abelson, 1992; Yu et al., 1995). For group II ribozymes, the catalytically important phosphate oxygens were similarly identified and suggest that domain V of group II introns has structure–function relationships analogous to the 3′ stem–loop of U6 in spliceosomes (Chanfreau and Jacquier, 1994). Correlations between pre-mRNA splicing and group II self-splicing are significant because the splicing mechanisms are quite similar in both systems and because the group II system is better understood and could serve as a model for RNA structure–function relationships in spliceosomes. In addition to backbone phosphate groups, specific 2′-hydroxyl groups of ribose rings play critical roles in binding specificity and/or chemical catalysis in self-splicing RNAs and ribozymes (Bevilacqua and Turner, 1991; Pyle and Cech, 1991; Musier-Forsyth and Schimmel, 1992; Perreault and Atlman, 1992; Pyle et al., 1992; Herschlag et al., 1993a, b; Smith and Pace, 1993; Strobel and Cech, 1993; Pley et al., 1994; Abramovitz et al., 1996). These results suggest that specific 2′-hydroxyl groups of U6 may play similar roles. In the present study, we identified which 2′-hydroxyl groups of U6 are required for the first or second step of splicing in a yeast in vitro splicing system (Lin et al., 1985). To screen all 2′-hydroxyls of yeast U6 at nucleotides 39–88, spanning the entire conserved central domain, synthetic U6 RNAs were constructed with deoxynucleotides incorporated site-specifically at single sites or at multiple contiguous sites. Each of the 2′-deoxy U6 RNAs was assayed for reconstitution of splicing activity in U6-depleted yeast extract (Fabrizio et al., 1989); thus the positions of important 2′-hydroxyl groups were revealed by assaying splicing defects for specific 2′-deoxy substituents. This study is the first to investigate deoxynucleotide substitutions in U6 snRNA. Results Synthesis of deoxynucleotide-substituted yeast U6 snRNAs To construct yeast U6 RNAs substituted with a 2′-deoxy- or 2′-O-methyl ribonucleotide at a selected site, each full-length RNA was prepared in pieces, annealed to a complementary single strand of DNA, and ligated using T4 DNA ligase (Kleppe et al., 1970; Moore and Sharp, 1992). Oligonucleotides for the central domain were chemically synthesized to incorporate a 2′-deoxy- or 2′-O-methyl ribonucleotide at selected sites from nucleotides 39 through 88. Initially, we prepared U6 RNA in three pieces, nucleotides 1–38, 39–59 and 60–112, such that the 5′ and 3′ end pieces were transcribed in vitro and the central piece (nucleotides 39–59) was synthesized chemically. The yield for the three-piece ligation was as high as 30%. Ligation yields are lower for incorrect 3′ and 5′ ends at a ligation junction, including the lack of a 5′-monophosphate. Indeed, our attempts to transcribe nucleotides 1–58 of U6 (from Sau3AI-cut plasmid) primarily produced transcripts that had extra, non-coded 3′-terminal nucleotides. Two improved strategies involved the total chemical synthesis of yeast U6 RNA in either four or five pieces. For the four-piece ligation, synthetic U6 RNA oligonucleotides 1–38, 39–59, 60–79 and 80–112 (all gel purified) were ligated to give full-length U6 in ∼20% yield. For the five-piece ligation, U6 RNA oligonucleotides 1–38, 39–59, 60–76, 77–94 and 95–112 were ligated in ∼12% yield. The four-piece ligation was used to incorporate a deoxynucleotide at each individual site from 60 through 79, and the five-piece ligation was used for each site from 80 through 88. We found that large quantities of RNA were best prepared via chemical synthesis of the entire RNA in segments for ligation. For the all-RNA pieces of U6, gel purification removed nearly all of the synthetic byproducts. To expedite preparation of the 2′-deoxy substituted RNA pieces (29 total), an improved method for automated RNA synthesis was followed (see Materials and methods). Synthetic 5′-O-trityl-protected oligonucleotides were purified on Nensorb columns to partially remove the aborted by-products of the chemical synthesis, and the eluted full-length 5′-OH oligonucleotides were phosphorylated and used for ligation. For a 0.2 μmol scale automated synthesis, the quantities of Nensorb-purified 18mer and 20mer products ranged from 1 to 30 nmol. For the five-piece ligations, we started with 40 pmol each of the all-RNA oligonucleotides and 60 pmol of the Nensorb-purified, deoxy-substituted oligonucleotide and obtained 2–3 pmol of gel–purified, full-length U6 RNA. In the four-piece ligations, the same amounts of the individual starting materials yielded 6–9 pmol of gel-purified full-length U6 RNAs. As less than 50 fmol of U6 RNA are needed per splicing assay in U6-depleted yeast extract, ample quantities were obtained to conduct the experiments described. Splicing activity of U6 RNAs substituted with multiple deoxynucleotides in the 5′ portion of the central domain Splicing assays were conducted in yeast extract using [α–32P]uridine-labeled actin pre-mRNA as the substrate. In order to assay U6 RNAs substituted with deoxynucleotides, the endogenous U6 snRNA in the yeast extract was depleted by oligo-directed RNase H digestion using a deoxyoligonucleotide (d1) complementary in sequence to nucleotides 28–54 of yeast U6 RNA (Fabrizio and Abelson, 1990). Over the period of incubation, oligonucleotide d1 is itself destroyed by endogenous DNases in the extract. In U6-depleted extract, each of the 2′-deoxy U6 RNAs was assayed for reconstitution of splicing activity (Fabrizio et al., 1989). As a control, wild-type U6 RNA was ligated from three pieces (see above), and the gel-purified, ligated U6 restored splicing activity in U6-depleted extract as efficiently as fully transcribed U6 (Figure 2, lanes 3 and 4). To screen the 2′-hydroxyl groups of U6 RNA at nucleotides 39–59 rapidly, we constructed three U6 RNAs substituted with seven contiguous deoxynucleotides at specific sites (deoxy box 39–45, 46–52 and 53–59) and assayed each for reconstitution of splicing activity (Figure 2, lanes 5–7). Only U6 RNA with deoxy box 39–45 (Figure 2, lane 5) reconstituted splicing activity, although not as efficiently as wild-type U6 RNA (Figure 2, lanes 3 and 4). Since deoxy box 39–45 was competent for splicing, we did not subsequently make 2′-deoxy substitutions at single sites for nucleotides 39–45. For deoxy boxes 46–52 and 53–59, the inability to reconstitute splicing activity (Figure 2, lanes 6 and 7) suggested that at least one 2′-OH group in this region is crucial for splicing or that multiple deoxynucleotide substitutions are inhibitory. Subsequently, we prepared U6 RNAs substituted with a single deoxynucleotide at each site from 46 through 59, and the singly substituted U6 RNAs were screened for reconstitution of splicing activity. Figure 2.Splicing activity of U6-depleted yeast extract reconstituted with U6 RNAs substituted with seven deoxynucleotides at specific contiguous sites. Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, then total RNA was isolated and separated on a denaturing polyacrylamide gel to assay splicing activity: mock-treated extract (lane 1); U6-depleted extract, i.e. extract treated with oligonucleotide d1 to digest endogenous U6 snRNA (lane 2); reconstitution of splicing activity with in vitro transcribed wild-type U6 snRNA (lane 3), wild-type U6 RNA ligated from three pieces (lane 4) and ligated U6 RNAs substituted with deoxynucleotides at nucleotides 39–45, 46–52 and 53–59 (lanes 5–7, respectively). Splicing substrate, reaction intermediates and products are labeled as follows: lariat intron–exon 2 (IVS-E2); lariat intron (IVS); pre-mRNA substrate (pre-mRNA); spliced mRNA product (mRNA). Download figure Download PowerPoint Splicing activity of U6 RNAs substituted with a single deoxynucleotide at each site from 46 through 88 Figure 3 shows results of splicing assays for U6 RNAs substituted with a single deoxynucleotide at each site from 46 to 59. For dA51-substituted U6, splicing activity was completely blocked (Figure 3, lane 12). Interestingly, splicing activity was not inhibited when A51 was replaced by a 2′-O-methyladenosine (Figure 3, lane 13). The 2′-O-methyl group cannot act as a hydrogen bond donor, nucleophile or coordinated anion (in contrast to 2′-OH groups), therefore 2′-O-methyl A51 does not function as such in spliceosomes. Nucleotide A51 is part of the functionally important ACAGAG sequence (Fabrizio and Abelson, 1990; Kandels-Lewis and Séraphin, 1993; Lesser and Guthrie, 1993). Figure 3.Splicing activity of U6-depleted yeast extract reconstituted with U6 RNAs substituted site-specifically with a single deoxynucleotide at each site from 46 through 59 (or with a 2′-O-methyl nucleotide at position 51). Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, then total RNA was isolated and separated on a denaturing polyacrylamide gel to assay splicing activity: mock-treated extract (lanes 1 and 8); U6-depleted extract (lanes 2 and 9); reconstitution of splicing activity with wild-type U6 RNA ligated from three pieces (lanes 3 and 10) and ligated U6 RNAs substituted at nucleotides 46–59 with a single deoxynucleotide (lanes 4–7 and 11–21, respectively). Download figure Download PowerPoint Surprisingly, single deoxynucleotide substitutions at positions 53–59 had no effect on splicing (Figure 3, lanes 15–21), despite the result that U6 with deoxy box 53–59 blocked splicing activity (cf. Figure 2). Figures 4 and 5 show results of splicing assays for U6 RNAs substituted with a single deoxynucleotide at each site from 60 to 88. In this region, one deoxynucleotide substitution, dA62 (Figure 4, lane 6), completely blocked splicing, and two others showed only traces of spliced products, dU70 (lane 14) and dC72 (lane 16, Figure 4). Nucleotide A62 is next to the functionally important CAGC sequence, part of U2–U6 helix I (Fabrizio and Abelson, 1990; Madhani and Guthrie, 1992), and nucleotides U70 and C72 are both part of the loop of the U6 intramolecular 3′ stem–loop (Fortner et al., 1994) (cf. Figure 1). As observed for a 2′-O-methyl substituent at nucleotide A51 (see above), a single 2′-O-methyl substituent at A62, U70 or C72 did not inhibit splicing activity (Figure 6), in contrast to the inhibitory effects of deoxynucleotide substitutions at these sites. Figure 4.Splicing activity of U6-depleted yeast extract reconstituted with U6 RNAs substituted site-specifically with a single deoxynucleotide at each site from 60 through 73. Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, then total RNA was isolated and separated on a denaturing polyacrylamide gel to assay splicing activity: mock-treated extract (lane 1); U6-depleted extract (lane 2); reconstitution of splicing activity with wild-type U6 RNA ligated from four pieces (lane 3) and ligated U6 RNAs substituted at nucleotides 60–73 with a single deoxynucleotide (lanes 4–17, respectively). Download figure Download PowerPoint Figure 5.Splicing activity of U6-depleted yeast extract reconstituted with U6 RNAs substituted site-specifically with a single deoxynucleotide at each site from 74 through 88. Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, then total RNA was isolated and separated on a denaturing polyacrylamide gel to assay splicing activity: actin pre-mRNA substrate (lane 1); mock-treated extract (lane 2); U6-depleted extract (lanes 3 and 4); reconstitution of splicing activity with in vitro transcribed wild-type U6 snRNA (lane 5) and ligated U6 RNAs (from five pieces) substituted at nucleotides 74–88 with a single deoxynucleotide (lanes 6–20, respectively). Download figure Download PowerPoint Figure 6.Splicing activity of U6-depleted yeast extract reconstituted with U6 RNAs substituted site-specifically with a single 2′-O-methyl nucleotide at position 62, 70 or 72. Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, then total RNA was isolated and separated on a denaturing polyacrylamide gel to assay splicing activity: mock-treated extract (lane 1); U6-depleted extract (lane 2); reconstitution of splicing activity with wild–type U6 RNA ligated from four pieces (lane 3) and ligated U6 RNAs substituted at positions 62, 70 and 72 with a single deoxynucleotide (lanes 4–6) or a single 2′-O-methyl nucleotide (lanes 7–9). Download figure Download PowerPoint Using a set of synthetic oligonucleotide stocks, we were able to reproduce our results. To verify that dA51- and dA62-substituted U6 completely blocked splicing activity and that dU70- and dC72-substituted U6 severely inhibited splicing, we prepared duplicate stocks of the deoxy-substituted oligonucleotides by resynthesizing and purifying them. Using the duplicate stocks, replication of the assay results for dA51-, dA62-, dU70- and dC72-substituted U6 RNAs (data not shown) confirmed that these individual deoxynucleotide substituents caused the splicing defects observed (cf. Figures 3 and 4). Native gel analysis of spliceosome assembly in yeast extracts reconstituted with deoxynucleotide–substituted U6 RNAs For the 2′-deoxy U6 RNAs that blocked splicing activity, it was important to distinguish whether a spliceosome assembly step was blocked or whether spliceosomes were fully assembled and the block occurred at the first catalytic step. Hence, U6 reconstitution experiments which blocked or inhibited splicing were repeated exactly as for the splicing assays except that product mixtures were split in order to assay both splicing activity (on a denaturing gel) and spliceosome assembly (on a native gel). A number of spliceosome assembly intermediates have been separated on native gels and characterized biochemically (Cheng and Abelson, 1987; Moore et al., 1993). In yeast, the biochemically characterized assembly intermediates are the ‘commitment complex’ containing pre-mRNA and U1 snRNP, the B complex (U2 snRNP addition), the A2-1 complex (U4–U5–U6 tri-snRNP addition) and the A1 complex (fully assembled spliceosomes). Complexes that contain first-step splicing products (A2-2) and second-step products (A2-3) were identified as well, but these are usually overlapped with A2-1 complexes on native gels. In the standard native gel assay for yeast spliceosome assembly, radiolabeled actin pre-mRNA is the assembly substrate, and partially and fully assembled spliceosomes migrate according to size with the fully assembled A1 complex at the top of the gel. The complexes are identified by inspection of the native gel with reference to a wild-type control lane. Because the A complexes are U6 dependent, they are readily identified via comparison with a U6-depleted control lane. To determine whether the samples generated A2-2 and A2-3 complexes (defined by the presence of spliced intermediates and products), the split portion of each sample was assayed for splicing activity on a denaturing gel. Figure 7 shows the spliceosome assembly assays for U6 RNAs substituted with a single deoxynucleotide at position 51, 61, 62, 68, 70, 71 or 72 (or with a 2′-O-methyl nucleotide at position 51). Split samples were assayed concurrently for reconstitution of splicing activity (data not shown). The mock-treated yeast extract assembled A and B complexes (Figure 7, lane 1), and the corresponding splicing assay showed the formation of lariat intermediate and mRNA product, indicating that the A2-2 and A2-3 complexes had formed in this sample. When endogenous U6 was depleted by oligo-directed RNase H digestion, no splicing occurred and no A complexes were assembled (Figure 7, lane 2), as expected for complexes that require U6 snRNP for assembly. When wild-type, ligated U6 RNA was added to the same stock of U6-depleted extract as a control, splicing activity and assembly of A complexes were restored (Figure 7, lane 3). Likewise in the same U6-depleted extract, the deoxynucleotide-substituted U6 RNAs which blocked splicing activity (i.e. dA51-, dA62-, dU70- and dC72-substituted U6), were all competent for assembly of A complexes (Figure 7, lanes 4–11). Although assembly of A complexes for dA62-substituted U6 was relatively diminished in the experiment shown (Figure 7, lane 7), replicate experiments (not shown) revealed that dA62 U6 assembled A complexes at relatively high levels. For dA51- and dA62-substituted U6, the absence of spliced intermediates and products on the corresponding denaturing gel (not shown) indicated that the A2-2 and A2-3 complexes were not produced in these samples. Figure 7.Spliceosome assembly of U6-depleted yeast extract reconstituted with U6 RNAs substituted site-specifically with a single deoxynucleotide at position 51, 61, 62, 68, 70, 71 or 72 (or with a 2′-O-methyl nucleotide at position 51). Samples were incubated with [α-32P]uridine-labeled actin pre-mRNA at 23°C for 30 min, treated with heparin, and separated on a native polyacrylamide gel: mock-treated extract (lane 1); U6-depleted extract (lane 2); reconstitution of spliceosome assembly with wild-type U6 RNA ligated from four pieces (lane 3) and ligated U6 RNAs substituted at the positions noted with a single deoxynucleotide (lanes 4–11). Download figure Download PowerPoint Concentration effects of deoxynucleotide-substituted U6 RNAs in U6-reconstituted yeast extract We tested various concentrations of U6 RNA in U6-reconstituted samples to determine whether the four inactive deoxy-substituted U6 RNAs (see above) were capable of reconstituting splicing activity at unusually high or low concentrations. Fabrizio et al. (1989) had titrated the concentration of wild-type U6 RNA for reconstitution of splicing activity in U6-depleted extract. As suggested by their titration data, the concentrations of added U6 RNA in our reconstituted samples were in the range of 15–20 nM. In addition to the four inhibitory deoxy-substituted U6 RNAs reported here, we tested the A51U mutant U6 RNA which blocks the second step of splicing and causes an accumulation of A2-2 complexes (Fabrizio and Abelson, 1990). As controls for normal assembly and splicing activity, we tested wild-type and dC66-substituted U6 RNAs. Reconstituted samples were split and assayed concurrently for splicing activity and spliceosome assembly (data not shown). For all reconstituted samples, the lowest concentration of added U6 (0.2 nM) produced the lowest levels of reconstituted activity, as expected from the published titration data. As the concentration of added U6 was increased in 10-fold increments (2, 20 and 200 nM), the levels of splicing activity increased for wild-type U6, dC66-substituted U6 and A51U U6 RNAs (this mutant blocked the second step of splicing as expected); however, no splicing activity was observed for dA51-, dA62-, dU70- and dC72-substituted U6 RNAs over the 103-fold range of concentrations. All of the reconstituted samples showed assembly of complete spliceosomes (A complexes) at increasing levels as the concentration of added U6 was increased, except at the highest level of added U6 (200 nM) which caused a shift from A and B complexes to poorly understood C complexes (Cheng and Abelson, 1987). This shift occurred for all U6 RNAs tested, and suggests that the 200 nM level of added U6 was high enough to assess the upper limits of U6 concentration effects. Therefore, the 15–20 nM concentration range for U6 reconstitutions was appropriate, and no unusual concentration effects were found for the four deoxy-substituted U6 RNAs that blocked or severely inhibited splicing activity. Discussion In this study of the essential functional groups of the U6 phosphate-ribose backbone, the 2′-OH groups of U6 ribose rings were substituted site-specifically with 2′-H groups to determine which 2′-hydroxyls of U6 are required for splicing in yeast extract. The 2′-hydroxyl group of RNA can contribute binding energy to a bonded interaction or can stabilize a transition state to catalyze a reaction (Bevilacqua and Turner, 1991; Pyle and Cech, 1991; Musier-Forsyth and Schimmel, 1992; Perreault and Atlman, 1992; Pyle et al., 1992; Herschlag et al., 1993a, b; Smith and Pace, 1993; Strobel and Cech, 1993; Pley et al., 1994; Abramovitz et al., 1996). In either case, it can act as a hydrogen bond donor or acceptor. In binding interactions, the 2′-hydroxyl group can be important for recognition of a ribonucleotide or a structural domain. In catalysis, it can act as an active site nucleophile, especially when deprotonated (e.g. by a metal hydroxide; Cech, 1987), and it can coordinate to active site metal ions that catalyze a reaction (Pley et al., 1994). Also, 2′-hydroxyls may help to organize water molecules for catalysis in an active site. In contrast, a 2′-H group cannot participate in hydrogen bonding or coordinate to metal ions. A deoxyribose ring in DNA conforms to a C2′ endo ring pucker, whereas a ribose ring in RNA conforms to a C3′ endo ring pucker. The single deoxyribonucleotide substitutions made in our experiments are not expected to alter the C3′ endo conformations of RNA structur