Portal de Periódicos da CAPES

Identification and characterization of a yeast homolog of U1 snRNP-specific protein C

1997; Springer Nature; Volume: 16; Issue: 13 Linguagem: Inglês

10.1093/emboj/16.13.4082

1460-2075

Jie Tang, Nadja Abovich, Margaret L. Fleming, Bertrand Séraphin, Michael Rosbash,

RNA modifications and cancer

Article1 July 1997free access Identification and characterization of a yeast homolog of U1 snRNP-specific protein C Jie Tang Jie Tang Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Nadja Abovich Nadja Abovich Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Margaret L. Fleming Margaret L. Fleming Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Bertrand Séraphin Bertrand Séraphin Gene Expression Program, EMBL, Meyerhofstrasse 1, D-69117 Heidelberg, Germany Search for more papers by this author Michael Rosbash Corresponding Author Michael Rosbash Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Jie Tang Jie Tang Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Nadja Abovich Nadja Abovich Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Margaret L. Fleming Margaret L. Fleming Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Bertrand Séraphin Bertrand Séraphin Gene Expression Program, EMBL, Meyerhofstrasse 1, D-69117 Heidelberg, Germany Search for more papers by this author Michael Rosbash Corresponding Author Michael Rosbash Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA Search for more papers by this author Author Information Jie Tang1, Nadja Abovich1, Margaret L. Fleming1, Bertrand Séraphin2 and Michael Rosbash 1 1Howard Hughes Medical Institute and Department of Biology, Brandeis University, Waltham, MA, 02254 USA 2Gene Expression Program, EMBL, Meyerhofstrasse 1, D-69117 Heidelberg, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:4082-4091https://doi.org/10.1093/emboj/16.13.4082 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info U1C is one of the three human U1 small nuclear ribonucleoprotein (snRNP)-specific proteins and is important for efficient complex formation between U1 snRNP and the pre-mRNA 5′ splice site. We identified a hypothetical open reading frame in Saccharomyces cerevisiae as the yeast homolog of the human U1C protein. The gene is essential, and its product, YU1C, is associated with U1 snRNP. YU1C depletion gives rise to normal levels of U1 snRNP and does not have any detectable effect on U1 snRNP assembly. YU1C depletion and YU1C ts mutants affect pre-mRNA splicing in vivo, and extracts from these strains form low levels of commitment complexes and spliceosomes in vitro. These experiments indicate a role for YU1C in snRNP function. Structure probing with RNases shows that only the U1 snRNA 5′ arm is hypersensitive to RNase I digestion when YU1C is depleted. Similar results were obtained with YU1C ts mutants, indicating that U1C contributes to a proper 5′ arm structure prior to its base pairing interaction with the pre-mRNA 5′ splice site. Introduction RNA splicing is a process that removes intervening sequences (introns) precisely and joins two flanking sequences (exons) together. Some highly structured large introns (group I and group II introns) can undergo self-splicing, i.e. the splicing reaction is catalyzed by the RNA itself (for reviews, see Saldanha et al., 1993; Michel and Ferst, 1995). In the case of nuclear pre-mRNA, splicing takes place in a large ribonucleoprotein complex called the spliceosome. Four small nuclear ribonucleoprotein (snRNP) particles, U1, U2, U5 and U4/U6, assemble into spliceosomes through an ordered, dynamic pathway. U1 snRNP recognizes the pre-mRNA 5′ splice site, U2 snRNP binds to the branch point sequence and U4/U5/U6 tri-snRNP enters the pathway subsequently to form an active spliceosome in which the splicing reactions take place (Rymond and Rosbash, 1988; Guthrie, 1991; Lamond, 1993; Madhani and Guthrie, 1994). As the major components of the spliceosome, each snRNP contains one or two small nuclear RNAs (snRNAs) and >10 proteins. A group of low molecular weight proteins (Sm proteins) is present in all snRNPs, and there are various numbers of specific proteins in each snRNP (Lührmann et al., 1990; Mattaj, 1993; Séraphin, 1995). The U1 snRNP in higher eukaryotic cells is made up of one U1 snRNA, eight Sm proteins and three U1-specific proteins. Several regions in U1 snRNA are functionally important and are highly conserved among different species. For instance, the 5′ arm of U1 snRNA base-pairs with the pre-mRNA 5′ splice site; stem–loops A and B are the binding sites of the U1snRNP-specific proteins U1-70K and U1A, respectively; and the Sm site, a single-stranded region near the 3′ end of the molecule, serves as the binding site for Sm proteins (Hamm et al., 1990; Lührmann et al., 1990). The U1-70K and U1A proteins bind to their target sequences via their N-terminal RNA-binding domains (RBDs). The U1-70K protein contains an arginine-rich region at its C-terminus, which probably interacts with SR proteins to facilitate the interaction between U1 snRNP and the pre-mRNA (Wu and Maniatis, 1993; Kohtz et al., 1994). There is an additional RBD at the C-terminus of U1A. This domain is not required for the binding of U1A to the U1 snRNA but may play a role in polyadenylation (Lutz and Alwine, 1994). Unlike U1-70K and U1A, the third U1-specific protein, U1C, does not contain an RBD, nor does it show high affinity binding to U1 snRNA. U1C carries a zinc finger-like structure at its N-terminus, which is required for the association of U1C with U1 snRNP (Nelissen et al., 1991). This association is, at least in part, mediated by protein–protein interactions (Nelissen et al., 1994). There is also evidence that U1C is required for a stable interaction between U1 snRNP and the pre-mRNA 5′ splice site (Heinrichs et al., 1990). Although the exact biochemical role of the U1C protein is unclear, it is the only U1 snRNP protein with a well-described functional role during in vitro splicing. The components of yeast (Saccharomyces cerevisiae) U1 snRNP have been partially characterized. Similarly to its higher eukaryotic counterpart, the yeast snRNP contains a single U1 snRNA molecule, a group of U1-specific proteins and a group of common snRNP proteins (Fabrizio et al., 1994). These common Sm proteins resemble their mammalian counterparts quite well (Roy et al., 1995; Séraphin, 1995). Yeast U1 snRNA is 5-fold larger than metazoan U1 snRNA, despite the conservation of most functionally important regions, e.g. the 5′ arm, stem–loop II and Sm site (Kretzner et al., 1987, 1990; Siliciano et al., 1987; Liao et al., 1990). The yeast snRNP also contains more than seven specific proteins, in contrast to three specific mammalian U1 snRNP proteins (Fabrizio et al., 1994). Two of these additional yeast U1 snRNP-specific proteins, PRP39p and PRP40p, have been cloned. Although they are essential for yeast U1 snRNP function, no homologs are known in higher eukaryotic cells (Lockhart and Rymond, 1994; Kao and Siliciano, 1996). Additional yeast U1 snRNP-specific proteins have been identified, but these results are not yet published (J.Tang, unpublished data). The yeast counterparts of two of the three mammalian U1 snRNP-specific proteins, U1-70K and U1A, have been identified and characterized. They resemble their metazoan counterparts in basic structure and U1 snRNA binding (Smith and Barrell, 1991; Kao and Siliciano, 1992; Liao et al., 1993; Tang and Rosbash, 1996). Yeast U1-70K contains one RBD, and the protein binds to the conserved stem–loop II in U1 snRNA (Smith and Barrell, 1991; Kao and Siliciano, 1992). Yeast U1A binds to a large loop on U1 snRNA stem III through its N-terminal RBD. Its C-terminal RBD also contributes to pre-mRNA splicing, but its precise function and binding site are unknown (Tang and Rosbash, 1996). The yeast counterpart of the third U1 snRNP-specific protein, U1C, has not yet been identified. Through database searching, we found an open reading frame (ORF) (as determined by the yeast genome sequencing project) that shares significant homology with the human U1C protein. By demonstrating that this protein is essential for splicing and is associated with U1 snRNP, we claim that it is the yeast version of the U1C protein. A combination of in vivo and in vitro assays indicate that the yeast protein contributes to U1 snRNP function like the contribution of U1C to mammalian U1 snRNP function. Moreover, the U1C-depleted snRNP and U1C ts mutants show a dramatic structural alteration at the 5′ end of snRNA. This is almost certainly related to the failure of the altered snRNPs to undergo a proper base-pairing interaction with the pre-mRNA 5′ splice site. Results Identification of a putative yeast U1C protein encoded by an essential gene Using the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990), we searched sequence databases with the human U1C protein sequence. The closest relative of human U1C in the yeast S.cerevisiae is a 231 amino acid (27 kDa) hypothetical protein named L8003.21 that we will refer to as YU1C in this manuscript. The sequence was generated by the yeast genome sequencing project. When YU1C was used as a query for a BLAST search, the human and Xenopus U1C proteins gave the highest scores. With a more sensitive search method (see Materials and methods), we also identified the U1C homologs in Caenorhabditis elegans and Schizosaccharomyces pombe, both of which were determined by systematic genome sequencing projects. A multiple sequence alignment of the human, Xenopus, C.elegans, S.pombe and S.cerevisiae proteins showed that the high homology region is limited to the first 38 residues (Figure 1). The human U1C protein has a zinc finger-like motif in this region, which is essential for its binding to U1 snRNP (Nelissen et al., 1991). The newly identified proteins (C.elegans, S.pombe and S.cerevisiae) are ∼50% identical to human U1C in this region, with the critical cysteine and histidine residues of the zinc finger-like sequence conserved. The C-terminal parts of these proteins are quite divergent, ranging from being proline rich in human and Xenopus, to lysine, serine and arginine rich in yeast. Figure 1.The alignment of the human, Xenopus, C.elegans, S.pombe and S.cerevisiae U1C proteins. This multiple sequence alignment was generated by the Clustal W program (Thompson et al., 1994), and edited by hand. The conserved residues among the five sequences are shaded in gray. Arrows point to the residues involved in the zinc finger-like structure. Download figure Download PowerPoint The gene encoding the YU1C protein (named YHC1 for yeast homolog of U1C) was cloned by PCR amplification of genomic yeast DNA with a pair of YHC1-specific primers. One copy of the endogenous YHC1 gene subsequently was deleted and replaced by the LEU2 gene by homologous recombination in a diploid strain. A successful disruption was determined by Southern blotting (data not shown). The heterozygous diploid was sporulated, and tetrads were dissected. All tetrads gave rise to at most two viable spores. All of the viable spores were LEU−, indicating that the disruption of the YHC1 gene was lethal (data not shown; see Materials and methods for details). Four spore tetrads were obtained from sporulation of the heterozygous diploids transformed with a URA3 plasmid carrying either the wild-type YHC1 or a hemagglutinin (HA)-tagged version under the control of the GAL promoter. In this case, all the LEU+ colonies were also URA+ and did not grow on 5-fluoro-orotic acid (5-FOA)-containing plates (data not shown). YU1C is associated with U1 snRNP and required for in vivo splicing To test whether the YU1C protein is associated with U1 snRNP, we tagged the YU1C protein with the HA epitope and immunoprecipitated yeast splicing extracts with the 12CA5 antibody (a monoclonal antibody recognizing the HA epitope). U1 snRNA is specifically co-immuno precipitated from YU1CHA extract (Figure 2; compare YU1CHA with the non-HA control). YU1C is more salt-sensitive than yeast U1A protein (Figure 2; compare YU1CHA with U1AHA), suggesting that YU1C may be a more loosely associated U1 snRNP protein. Figure 2.The HA-tagged YU1C protein co-immunoprecipitates with U1 snRNA. Splicing extracts from the indicated strains were incubated with Gamma-Bind beads pre-coated with 12CA5 antibody. The beads were washed with increasing salt. RNA co-immunoprecipitated on beads was recovered and reverse transcribed with U1 and U2 snRNA-specific primers. U1 and U2 snRNA are indicated by arrows. The KCl concentrations in the washing buffer are indicated above. Lanes 1–3, extract with tagged U1A (U1AHA); lanes 4–6, extract without tagged protein (non-HA); lanes 7–9, extract with tagged YU1C (YU1CHA); lanes 10–12, 10% of the supernatant from the immunoprecipitation reactions indicated above. Download figure Download PowerPoint In order to conditionally deplete the yeast U1C protein in vivo, we put the YHC1 coding sequences under a GAL-controlled promoter (GAL-YHC1). A strain that carried the GAL-YHC1 plasmid as the only source of YU1C protein grew well in galactose-containing medium and somewhat more slowly in glucose-containing medium. The ΔYHC1/GAL-YHC1 strain grown in glucose was harvested at several time points and total yeast RNA was extracted. Pre-U3 and U3 snRNA were detected by primer extension. The pre-U3 level increased dramatically during growth in glucose (Figure 3A), indicating that in vivo splicing was significantly impaired by depletion of the yeast U1C protein. Figure 3.In vivo splicing in the yeast U1C-depleted strain. (A) The ΔYHC1/GAL-YHC1 strain was grown in galactose-containing medium (lane 1) for 10 h or in glucose-containing medium for 4 (lane 2), 7 (lane 3) or 10 h (lane 4). Total RNA was extracted from each culture and the U3 snRNA level was assayed by reverse transcription with the U3-specific primer DT1967. Pre-U3A and pre-U3B as well as U3 snRNA are indicated by arrows. The reverse transcription of U2 snRNA served as an internal loading control. (B) Same as (A), except the U1 snRNA level was assayed by primer extension with U1 snRNA-specific primer DT586. The U1 and U2 snRNA bands are indicated by arrows. Download figure Download PowerPoint Lack of an effect of YU1C on U1 snRNP assembly U1 snRNA levels were monitored during glucose depletion (Figure 3B). Remarkably, there was no detectable decrease. Because the cells grow normally during the first few doublings in glucose, the lack of an effect on U1 snRNA levels indicates that U1 snRNP accumulation and stability are largely unaffected by the absence of the YU1C protein. Several additional approaches were applied to assess the effect of YU1C on snRNP assembly. An in vivo dimethylsulfate (DMS) modification assay indicates that the recently identified U1A-binding site is protected in the YU1C-depleted strain (Tang and Rosbash, 1996; Figure 4A, lane 8). Lane 4 shows the pattern in this region when the U1A protein is absent. In vitro, an anti-70K antibody co-immunoprecipitates U1 snRNA from a YU1C-depleted extract as efficiently as from a wild-type extract (Figure 4B, compare lanes 6 and 9). Both assays indicate that other U1 snRNP-specific proteins assemble into U1 snRNP in the absence of the YU1C protein, consistent with the lack of a detectable effect of YU1C depletion on U1 snRNP accumulation and stability. Figure 4.YU1C-depleted snRNP retains other specific proteins. (A) U1A protein protects its binding site in the YU1C-depleted snRNP. Twenty-five ml cultures of the wild-type strain (lanes 1 and 2), the ΔU1A strain (lanes 3 and 4) and the ΔYHC1/GAL-YHC1 strain [grown in galactose- (lanes 5 and 6] or glucose- (lanes 7 and 8) containing medium] were treated with 200 μl of 1:2 diluted DMS. Total yeast RNA was extracted from DMS-treated strains and DMS modification patterns were assayed by reverse transcription with U1-specific primer DT2502. Lanes 1, 3, 5 and 7, negative controls without DMS treatment; lanes 2, 4, 6 and 8, DMS-treated samples. Lanes 9 and 10, U1 snRNA sequence ladders. U1 snRNA loop IIIc, the U1A protein-binding site, is indicated by the arrow. (B) U1-70K protein associates with YU1C-depleted U1 snRNP. Splicing extracts from the ΔSNP1/U1–70KHA strain (lanes 1–3) or from the ΔYHC1/GAL-YHC1 strain grown in galactose- (lanes 4–6) or glucose- (lanes 7–9) containing medium are immunoprecipitated by anti-70K antibody. The U1 snRNA co-immunoprecipitated with U1-70K protein was recovered from beads and assayed by reverse transcription with 32P-labeled U1- and U2-specific primers. The 12CA5 antibody against the HA epitope was used as a positive control (lane 2) and the pre-immune serum was used as a negative control (lanes 5 and 8); 10% of supernatant is shown to indicate immunoprecipitation efficiency (lanes 1, 4 and 7). The U1 and U2 snRNA bands are indicated by arrows. Download figure Download PowerPoint The yeast U1C protein is important for the U1 snRNP–pre-mRNA interaction in vitro To verify that YU1C depletion affects U1 snRNP function, we made splicing extracts from the ΔYHC1/GAL-YHC1 strain grown in galactose- or glucose-containing medium and compared their ability to form commitment complexes and spliceosomes with pre-mRNA. The extracts from the glucose-grown strain formed much less commitment complex and spliceosomes than the extracts from the galactose-grown strain (Figure 5A), indicating that the YU1C-depleted U1 snRNP is unable to form a stable pre-mRNA complex. Consistent with the commitment complex phenotype, the glucose-grown extract also failed to support substantial levels of in vitro splicing (Figure 5B). Figure 5.Yeast YU1C-depleted extracts form less commitment complexes and spliceosomes. (A) Splicing extracts were made from the ΔYHC1/GAL-YHC1 strain grown in either galactose-containing medium for 10 (lanes 1 and 5) or 20 h (lanes 2 and 6) or glucose-containing medium for 10 (lanes 3 and 7) or 20 h (lanes 4 and 8). 32P-Labeled pre-mRNA was incubated with different extracts under splicing conditions with (lanes 1–4) or without (lanes 5–8) oligonucleotide-mediated RNase H digestion of U2 snRNA, and the complexes were analyzed by native gel electrophoresis. The positions of commitment complexes (CC) and spliceosome (SP) are indicated. (B) The same reactions from lanes 5–8 of (A) were phenol–chloroform extracted and ethanol precipitated. The 32P-labeled pre-mRNA was assayed on a 15% denaturing polyacrylamide gel. The positions of the lariat intermediate, the lariat intron product and the pre-mRNA are indicated by arrows. Download figure Download PowerPoint In another in vitro assay, a biotinylated pre-mRNA substrate was pre-bound to streptavidin beads and then incubated with YU1C-depleted or wild-type extracts under splicing conditions. The U1 snRNA associated with the substrate was assayed by reverse transcription with a U1-specific primer. In the depleted extract, much less U1 snRNA was associated with the pre-mRNA (Figure 6A). To verify that the base-pairing potential of the 5′ arm was affected by YU1C depletion, we substituted the pre-mRNA substrate in the previous experiment with a biotinylated 2′-O-methyl RNA oligonucleotide complementary to the U1 snRNA 5′ arm (Figure 6B). A similar result was obtained, i.e., the YU1C-depleted snRNP binds poorly to the oligonucleotide (Figure 6B). Taken together, these in vitro results suggest that the YU1C protein potentiates the base-pairing interaction between the 5′ arm of U1 snRNA and a 5′ splice site sequence. Figure 6.YU1C-depleted U1 snRNP associates with pre-mRNA less efficiently. (A) Extracts from strain ΔYHC1/GAL-YHC1 grown in galactose- (lanes 1 and 3) or glucose- (lanes 2 and 4) containing medium were incubated with biotinylated pre-mRNA (pre-bound to streptavidin beads) under splicing conditions. U1 snRNA associated with pre-mRNA was assayed by reverse transcription with U1-specific primer DT586. Lanes 1 and 2, 10% of input; lanes 3 and 4, U1 snRNA bound to pre-mRNA. The U1 snRNA is indicated by an arrow. (B) Same as (A), except a biotinylated 2′-O-methyl RNA oligonucleotide complementary to the U1 snRNA 5′ arm region was used instead of pre-mRNA. Download figure Download PowerPoint Effects of YU1C on U1 snRNP structure Although glucose depletion of YU1C protein does not grossly affect the accumulation or stability of U1 snRNP (Figure 3B), we searched for a more subtle effect on U1 snRNP structure that might accompany the functional deficit of the YU1C-depleted extracts. U1 snRNP was digested with several RNases, to see if the missing YU1C causes any detectable change in U1 snRNA sensitivity. RNase V1, RNase A and nuclease S1 partial digestions did not detect any difference between the YU1C-depleted and wild-type extracts (data not shown). This is consistent with the presence of the U1A and the U1-70K proteins. However, RNase I probing showed cleavage sites, specific for the YU1C-depleted extract (Figure 7A). These two RNase I-sensitive sites reside in the U1 snRNA 5′ arm (Figure 7B), at the CC sequence that base-pairs with the highly conserved GG at the 5′ splice site. Figure 7.U1 snRNP structure probing with RNase I. (A) Extracts from strain ΔYHC1/GAL-YHC1 grown in galactose- (lanes 1 and 2) or glucose- (lanes 3 and 4) containing medium were partially digested with RNase I. The U1 snRNA digestion pattern was assayed by reverse transcription of total yeast RNA with a U1-specific primer. (−) control without RNase I; (+) RNase I partial digestion. Lanes 5 and 6, U1 snRNA sequence ladders (an in vitro transcribed U1 snRNA, which contains one extra nucleotide at its 5′ end, is used as a template). RNase I-hypersensitive sites are indicated by arrows. (B) U1 snRNA secondary structure. The RNase I-hypersensitive sites in the YU1C-depleted extract are mapped on the 5′ arm of U1 (indicated by arrows). Download figure Download PowerPoint To verify and extend this observation, we isolated several YU1C temperature-sensitive mutants (see Materials and methods). Three YU1C ts mutants that grow well at 25 but not at 37°C were characterized in some detail. First, the splicing efficiency of the U3 snRNA was assayed in these ts strains. Even at 25°C, these strains accumulated much more pre-U3B RNA than the wild-type control strain (Figure 8 and data not shown), indicating that in vivo splicing of these ts strains is already compromised at the permissive temperature. After switching the ts strains to 37°C for 12 h, the pre-U3B RNA levels were unaffected but U3 snRNA levels were dramatically decreased (Figure 8; to allow a comparison of U3 snRNA levels, the gel was dramatically underexposed compared with Figure 3A). In the wild-type control strain, U3 RNA levels were indistinguishable between 25 and 37°C (Figure 8, compare lanes 7 and 8). The result shows that the three YU1C ts mutants are already defective in splicing at 25°C, which becomes even more prominent after a shift to 37°C. We then made extracts from these three ts strains, incubated either at permissive or non-permissive temperature. In all cases, no complexes or splicing were observed (data not shown). Partial digestion of the 25°C extracts with RNase I indicated the hypersensitive cleavage sites within the U1 snRNA 5′ arm (Figure 9, lane 4; data not shown), consistent with the observation that the ts strains are compromised in splicing even at the permissive temperature. For the ts strains shifted to 37°C, the cleavages are stronger than those at 25°C (compare lane 6 with lane 4; and data not shown). The observations on the ts mutants confirm the glucose depletion data and link the presence or function of the YU1C protein to the structure of the U1 snRNA 5′ arm. Figure 8.In vivo splicing phenotype of YU1C ts strains. YU1C ts strains and a wild-type control strain were incubated at 25 or 37°C for 12 h. Total yeast RNA was extracted, and U3 snRNA levels were assayed by reverse transcription with a U3-specific primer. A U2 snRNA-specific primer was mixed with the U3 primer as an internal loading control. Lanes 1 and 2, YU1C ts2; lanes 3 and 4, YU1C ts3; lanes 5 and 6, YU1C ts4; lanes 7 and 8, wild-type control. Lanes 1, 3, 5 and 7, strains incubated at 25°C; lanes 2, 4, 6 and 8, strains incubated at 37°C. The pre-U3B, U2 snRNA and U3 snRNA bands are indicated by arrows. The pre-U3A bands are too faint at this level of exposure. Download figure Download PowerPoint Figure 9.RNase I structure probing of YU1C ts strains. YU1C ts or wild-type strains were incubated at 25 or 37°C for 12 h. Splicing extracts were made from those strains and subjected to RNase I partial digestion. Digestion patterns were assayed by reverse transcription with a U1 snRNA-specific primer. (−) control without RNase I; (+) RNase I partial digestion. Lanes 1 and 2, RNA sequence ladders; lanes 3 and 4, YU1C ts2 strain at 25°C; lanes 5 and 6, YU1C ts2 strain at 37°C; lanes 7 and 8, wild-type strain at 25°C; lanes 9 and 10, wild-type strain at 37°C. RNase I-hypersensitive sites are indicated by arrows. YU1C ts3 and ts4 strains show similar phenotypes (data not shown). Download figure Download PowerPoint Discussion The results of this study indicate that the completely conserved 10 nucleotides at the 5′ end of U1 snRNA have an RNP structure that is important for the base-pairing interaction with a 5′ splice site. There are few other cases in which a missing or mutant snRNP protein has been associated with an alteration in snRNP structure (Tang and Rosbash, 1996). Moreover, there are no other cases of which we are aware in which a change in snRNP structure is paralleled by a change in in vitro snRNP function. Although the binding results (Figures 5 and 6) are consistent with those from previous in vitro studies that demonstrated a comparable lack of activity of U1C-depleted mammalian snRNPs (Heinrichs et al., 1990), an effect on snRNP structure was not predictable. For example, the U1 snRNP–pre-mRNA base-pairing might be stabilized by one or more U1 snRNP proteins without any detectable impact of these proteins on snRNP structure prior to the snRNP–substrate interaction. Enhanced nuclease sensitivity implies that access to the 5′ arm is normally blocked by an interaction with one or more snRNP proteins. A further implication is that this interaction positively contributes to the base-pairing between the U1 snRNP 5′ arm and the 5′ splice site. Either YU1C interacts directly with the 5′ arm and subserves this role, or YU1C plays a more indirect role and affects the direct interaction between the 5′ arm and other U1 snRNP proteins. Biochemical approaches will be necessary to identify the proteins that interact directly with the 5′ arm. The conservation between the yeast and human U1C protein is limited to the common N-terminal zinc finger-like region (C2H2 motif), ∼38 amino acids in length. Thus it was not obvious that the identified ORF would encode the yeast U1C equivalent. However, it has been shown that the first 40 residues of human U1C are necessary and sufficient for U1C–U1 snRNP association (Nelissen et al., 1991), and this region is also required for U1C homodimerization (Gunnewiek et al., 1995). The high degree of phylogenetic conservation of the C2H2 motif is consistent with the idea that this motif is very important for U1C function. However, a truncated yeast U1C protein carrying the first 80 residues fails to rescue the lethality of the ΔYHC1 strain (data not shown), indicating that the C2H2 motif may not be sufficient for the function of the protein. The C2H2 motif also exists in other splicing factors such as Prp6p, Prp11p and Prp9p (Legrain and Choulika, 1990). Although the C2H2 motifs are not required for the protein–protein interactions within the Prp9p–Prp11p–Prp21p complex, the deletion of one or two motifs from Prp9p or Prp11p is dominant lethal, suggesting a functional role for this kind of motif (Legrain and Chapon, 1993; Legrain et al., 1993). Whether this zinc finger-like motif binds to zinc, to RNA (the 5′ arm?) or to other proteins remains unknown. The C-terminal part of the yeast U1C protein is rich in lysine, arginine and serine. A similar feature can also be found in the C-terminus of the yeast U1-70K protein. The function of this type of sequence motif has not been determined. It may resemble the function of the arginine/serine-rich domain in human U1-70K and other metazoan splicing factors, which is involved in protein–protein interactions (Wu and Maniatis, 1993; Kohtz et al., 1994). It could also facilitate protein–RNA interactions via its positively charged residues. Alternatively, it may only act indirectly, by aiding the maintenance of a proper YU1C protein conformation. The corresponding region of human U1C is proline rich and has not been assigned a functional role. The co-immunoprecipitation of U1 snRNA with HA-tagged U1C protein (YU1CHA) is salt sensitive, i.e. U1 snRNA co-immunoprecipitation decreases substantially when the salt concentration is >150 mM (Figure 2). The HA-tagged yeast U1-7