|Home | About | Journals | Submit | Contact Us | Français|
Alternative splicing makes a major contribution to proteomic diversity in higher eukaryotes with ~70% of genes encoding two or more isoforms. In most cases, the molecular mechanisms responsible for splice site choice remain poorly understood. Here, we used a randomization-selection approach in vitro to identify sequence elements that could silence a proximal strong 5′ splice site located downstream of a weakened 5′ splice site. We recovered two exonic and four intronic motifs that effectively silenced the proximal 5′ splice site both in vitro and in vivo. Surprisingly, silencing was only observed in the presence of the competing upstream 5′ splice site. Biochemical evidence strongly suggests that the silencing motifs function by altering the U1 snRNP/5′ splice site complex in a manner that impairs commitment to specific splice site pairing. The data indicate that perturbations of non-rate limiting step(s) in splicing can lead to dramatic shifts in splice site choice.
In higher eukaryotes, the majority of pre-mRNAs are subject to alternative splicing, a process that can be regulated according to developmental stage, cell type or in response to signal transduction pathways (reviewed in Black, 2003; Blencowe, 2006; House and Lynch, 2008). Splicing patterns can be remarkably complex, with some pre-mRNAs processed to yield dozens or even thousands of distinct isoforms (reviewed in Black and Graveley, 2006).
Intensive bioinformatic and experimental analyses have begun to identify specific sequence elements that either positively (splicing enhancers) or negatively (splicing silencers) influence splice site choice and, in many cases, specific trans-acting factors which recognize these elements have been characterized (reviewed in Black and Graveley, 2006; Blencowe, 2006; Wang and Burge, 2008).
With regard to silencers, there is evidence in support of several distinct mechanisms by which these elements exert their inhibitory effects. One straightforward mechanism is “bind and block” wherein a protein factor binds to a silencing element and sterically prevents the binding of a splicing factor (e.g. Kanopka et al., 1996; Mayeda and Krainer, 1992; Merendino et al., 1999; Shin et al., 2004; Ule, et al., 2006; Valcarcel et al., 1993; Wagner and Garcia-Blanco, 2001; Zheng, et al., 1998, Zhu et al., 2001). A second mechanism is silencer-promoted formation of nonfunctional or “dead end” complexes (e.g. Agris et al., 1989; Kan and Green, 1999; Giles and Beemon, 2005; House and Lynch, 2006; Labourier et al., 2001). Such complexes apparently contain all of the factors necessary for splicing but are unable to execute the reaction, presumably because a crucial conformation or conformational change is blocked. A third mechanism is blockage of communication between splice sites either by looping out of the affected site (e.g. Blanchette and Chabot, 1999) or binding of a repressive complex downstream of the regulated site (e.g. Nagengast et al., 2003; Sharma et al., 2005; Sharma et al., 2008). In these cases, splice site recognition does not appear to be affected, but productive association of 5′ and 3′ splice sites is prevented by mechanisms which have not yet been elucidated. There are also numerous examples of silencers which have been shown to bind specific trans-acting factors; but how these proteins exert their negative effects is largely unknown (reviewed in Black and Graveley, 2006; Fu, 2004; Hastings and Krainer, 2001; Matlin et al., 2005).
While the mechanisms of splicing silencers are beginning to be elucidated in the context of regulated exons, it is not clear whether similar elements or mechanisms are operative in the repression of “splice sites” that are never used in splicing. In this regard, it is well established that potential 5′ splice sites (including those that perfectly match the consensus recognition site for U1 snRNP) far outnumber authentic 5′ splice sites (Senapathy et al., 1990; Sun and Chasin, 2000). Furthermore, it is not clear why “pseudoexons”, exon-sized sequences that are bounded by sequences indistinguishable from functional 3′ and 5′ splice sites, are ignored by the splicing machinery (Cote et al., 2001; Sun and Chasin, 2000).
To gain further insight into potential molecular mechanisms by which splice sites are silenced, we employed a randomization-selection (SELEX) (Tuerk and Gold, 1990) strategy in vitro designed to identify all possible sequences at specific intronic and exonic positions that could silence a consensus 5′ splice site. Because the experimental design demanded that a “perfect” proximal 5′ splice site be silenced in the presence of a weakened distal 5′ splice site, we anticipated that we would recover elements that would inactivate or occlude the proximal site through formation of stable protein-RNA complexes. Focusing only on elements that conferred the strongest silencing phenotypes, we identified two exonic and four intronic motifs whose presence caused nearly complete inhibition of proximal splicing and concomitant activation of the weak upstream site. Remarkably, none of these motifs functioned by sequestering or inactivating the strong 5′ splice site. Rather, in the absence of the upstream 5′ splice site, the consensus site remained fully functional. These results demonstrate that kinetic effects on non-rate limiting steps can elicit dramatic differences in splicing patterns and may help to explain alternative splicing phenotypes observed when the levels of many splicing factors including basal components of the spliceosome (e.g. Karni et al., 2007; Olson et al., 2007; Park et al., 2004) are even modestly altered.
To identify potential splicing regulatory signals that could silence a strong 5′ splice site, we generated a synthetic pre-mRNA containing two alternative 5′ splice sites; a weak upstream 5′ splice site and a strong proximal 5′ splice site (see Figure 1 and Experimental Procedures). In in vitro splicing assays, the strong proximal 5′ splice site was used almost exclusively (Figure 1A). Importantly, the weak distal 5′ splice site remained functional since inactivation of the proximal 5′ splice site by mutation resulted in distal splicing (Figure 1A). We used this base construct to generate pre-mRNAs containing completely randomized regions either upstream or downstream of the strong 5′ splice site. The pool randomized at positions +11 to +22 relative to the proximal site was used to identify intronic silencers, while the pool randomized at the positions −18 to −7 was used to identify exonic silencers (see Experimental Procedures). Twelve nucleotides were chosen for randomization because this length represents at least two fold coverage of the binding site size (6 nt) of most RNA binding proteins (see Fairbrother et al., 2002) and the total pool (~107 variants) can be easily accommodated in standard in vitro splicing reactions (3 ng of RNA contain ~1010 molecules of substrate). The positions for insertion of the randomized sequences were chosen so as not to overlap with the minimal binding site of U1 snRNP; [−6 to +10 relative to the splice junction (P.A.M., J.A.D. and T.W.N unpublished; Mount et al., 1983)]. To evaluate the quality of each pool, in vitro transcribed RNAs were subjected to primer extension sequencing; equal distribution of all bases in the randomized regions indicated no sequence bias (data not shown and also see Figure S4A). To facilitate the selection (see below), the 3′ splice site was mutated from AG/G to UC/C such that splicing was arrested after the first transesterification reaction (e.g. Reed, 1989), resulting in accumulation of lariat-3′ exon intermediates. This 3′ splice site mutation did not affect 5′ splice site choice (compare Figure 1A with 1B). When the pools were assayed for splicing in vitro, they spliced identically to the unsubstituted RNAs, i.e. the proximal site was used almost exclusively; indicating that most sequences within the pool had no effect on splice site choice (Figure 1B, compare lanes 1 and 3; Figure S4B).
To identify those sequences which could modulate use of the proximal site, we used the selection strategy outlined in Figure 1C. In brief, the pools of templates were transcribed in vitro and because we considered it likely that the most common “silencing elements” in the pool would be sequences that could base pair with the 5′ splice site and thus occlude it through formation of stable secondary structure, we introduced a step to exclude such elements prior to splicing assays. Because purified U1 snRNP binds stably to consensus 5′ splice sites (the strong site) but not to splice sites mutated at the +5+6 positions (the weak site) (P.A.M., J.A.D., T.W.N., unpublished), we could use U1 snRNP binding to enrich for RNA molecules in which the strong proximal 5′ splice site remained accessible. Accordingly, we mixed purified U1 snRNP (Hochleitner et al., 2005) with the starting pools and recovered bound RNAs using anti-U1A antisera (see Experimental Procedures). Recovered RNAs were then deproteinized and used as substrates for in vitro splicing in HeLa cell nuclear extract.
After splicing, the RNAs were separated on denaturing polyacrylamide gels and the region containing distal lariat intermediates was excised from the gel. Following debranching (see Experimental Procedures), the linear RNA molecules containing the proximal 5′ splice site and surrounding sequence were amplified by RT-PCR. These molecules were then used to regenerate the starting constructs by overlapping PCR (Figure 1C).
As shown in Figure 1D (lane 0), there was negligible distal splicing in the starting pools but the proportion of distal splicing intermediates rose rapidly after iterative rounds of selection; maximum accumulation of distal intermediates was reached after only four rounds of selection for the intronic position (Figure 1D, lanes 1–4) and seven rounds for the exonic position (data not shown). At this point, the populations were recovered and intact templates generated as described above. These pools of DNAs were then cloned and propagated as libraries and pre-mRNAs transcribed from individual clones were analyzed for their splicing behavior. Although there was clone to clone variance, the majority demonstrated a pronounced preference for the distal splice site (data not shown and see Figure 2B). By this type of analysis, we recovered 106 clones containing intronic silencers and 52 clones containing exonic silencers that displayed predominant use of the distal 5′ splice site; clones which demonstrated substantive but less dramatic shifts in splice site choice were not analyzed further.
All clones that showed dramatic splicing phenotypes were sequenced. The 89 unique intronic silencers (12-mers) and the 47 unique exonic silencers (12-mers), together with 2-nt flanking regions were then hierarchically clustered to extract groups of similar sequences. Four intronic motifs and two exonic motifs were identified (see Table S1 for a complete list of the sequences). The 6 distinct motifs are presented as logos in Figure 2A and the splicing behavior of a specific representative of each class is shown in Figure 2B. Importantly, none of the silencing motifs markedly reduced overall splicing efficiency as a reciprocal relationship was observed between the reduction in proximal splicing and enhancement of distal splicing.
To assess the general significance of the selected silencer elements to splicing of human pre-mRNAs, they were compared to three sets of previously characterized splicing silencer motifs (Wang et al., 2004; Wang et al., 2006; Zhang et al., 2003; Zhang and Chasin, 2004). In particular, Motifs A, C, E and F, but not B or D, showed highly significant similarities (for details, see Figure S1A and Table S2). The potential relevance of the silencer motifs was then examined further by determining their occurrence on a genome-wide scale. We found that intronic Motifs A and C are highly enriched downstream of pseudo 5′ splice sites relative to constitutive 5′ splice sites (P<10−121 and P<10−74); the exonic Motifs E and F were 3 to 4 times more abundant in pseudo exons compared to real exons (P<10−7, for details, see Figure S1B). These results are consistent with a role for several of the identified motifs in repressing pseudo 5′ splice sites.
Because the informatic analyses indicated that the silencing elements identified in vitro were highly likely to be relevant in vivo, specific representatives of each class, in the same context as that assayed in vitro, were introduced into a mammalian expression vector and transfected into HeLa cells; the functional 3′ splice site (AG/G) was used in vivo. As shown in Figure 3A, the splicing behavior of the base construct was identical to that observed in vitro; i.e. nearly exclusive use of the proximal strong 5′ splice site in the wt construct (Figure 3A, lane 5) and exclusive use of the distal site when the proximal site was inactivated by mutation (Figure 3A, lane 8). Importantly, all of the silencers increased distal splicing (Figure 3A, lanes 1–4, 6, 7). The most dramatic effect was observed with exonic Motif E (Figure 3A, lane 7); the in vivo splicing phenotypes of exonic Motif F and the four intronic elements were more modest than those observed in vitro. In addition, random sequences other than the selected silencers were tested and they did not alter the 5′ splice site choice (Figure S4B).
While reduction in the magnitude of regulation in vivo could result from several potential variables, we hypothesized that the preference for use of the proximal splice site might be stronger in vivo than in vitro. In this case, weakening of the proximal site would be predicted to maintain the preferential use of the proximal site and might allow more effective silencing. Indeed, when we weakened the proximal site to make it identical to the distal site, proximal splicing predominated (Figure 3C, lane 5). Strikingly, the potency of all of the silencing elements was greatly enhanced under these conditions (Figure 3C, lanes 1–4, 6, 7); all now caused an almost complete shift to distal splicing.
If the silencing elements functioned by occluding or otherwise inactivating the affected 5′ splice site, it would be predicted that overall splicing would be drastically impaired when the elements were present at both 5′ splice sites in the dual splice site construct. However, this expected behavior was not observed. Remarkably, when the silencing elements were inserted at both 5′ splice sites, the wild type pattern of splicing was restored (Figure 3B and 3D). Importantly, there was no substantive reduction in the overall extent of splicing; quantitation showed that the overall splicing efficiency of the dual silencer constructs ranged from 83% (Motif E) to 100% (Motif C) of that measured for the construct lacking silencer elements.
We then asked if similar patterns could be observed in vitro. As shown in Figure 3E, this was indeed the case; in the absence of silencing elements, there was exclusive use of the proximal site, and when silencing elements were placed at both sites the wild type pattern was restored. As was the case in vivo, the presence of silencing elements at both sites did not markedly affect the overall level of splicing; quantitation showed that the overall splicing efficiency of the dual silencer constructs ranged from 70% (Motif E) to 92% (Motif C) of that measured for the pre-mRNA lacking silencer elements. An important and unexpected conclusion that emerged from both the in vivo and in vitro experiments was that the silencers did not function by preventing use of the affected site, but rather changed in some way the ability of the affected site to compete with the unsilenced site (see Discussion).
The results described above predicted that the silencing elements would have little if any effect on splicing in single 5′ splice site constructs (i.e. in the absence of a competing 5′ splice site). As shown in Figure 3F, this prediction was borne out by experiment; when the upstream 5′ splice was removed, none of the silencing elements had a pronounced effect on the accumulation of spliced product.
Because the in vivo results indicated that the silencer elements were more potent when the affected 5′ splice site was weakened (see Figure 3C), it was of interest to determine if this increased activity would translate into observable kinetic effects on splicing rate. Accordingly, we carried out time courses of splicing using single 5′ splice site constructs where the site was strong or weak with or without each of the silencing elements; we also measured the rate of splicing of each of these constructs in the presence of a strong 3′ splice site or when the 3′ splice site was weakened by purine substitutions into the polypyrimidine tract (Tian and Maniatis, 1994)(See Figure 4A). Figure 4B shows the results of the 60 minute time point for the panel of constructs with two of the silencing elements (Motifs D and E); Figure 4C–F shows the time course of splicing for all of the constructs and Figure 4G summarizes the data obtained at the 40 minute time points.
The data reveal several interesting points. First, there were perceptible but subtle decreases in splicing rate in the presence of any of silencing elements when both splice sites were strong (Figure 4C); essentially the same kinetics were observed when only the 3′ splice site was weakened (Figure 4D). Second, when the 5′ splice site was weakened in the presence of the strong 3′ splice site, there was a marked decrease in splicing rate observed with four of the silencing elements (D, B, A, and E); this effect was not observed with Motifs C and F (Figure 4E). Third, weakening of the 3′ splice site amplified the kinetic effects of the four intronic silencing elements, but had little if any effect on the two exonic silencers (Figure 4F). Because recognition of the weakened 3′ splice site relies on participation of the 5′ splice site (e.g. Barabino et al., 1990), the simplest explanation for these results is that the intronic silencers make a weak 5′ splice site less able to facilitate loading of factors at the 3′ splice site. Taken as a whole, the kinetic data strongly suggest that the silencing elements exert their effects through modulation of the affected 5′ splice site and when that site is strong, the effects on splicing are essentially invisible.
Because recognition at the 5′ splice site by U1 snRNP is almost surely rate limiting for splicing in general (e.g. Grabowski et al., 1985; Jamison et al., 1992; Seraphin and Rosbash, 1989), it seemed unlikely that the silencing elements could function by affecting this step of the reaction (Reed and Maniatis, 1986). Indeed, when we measured U1 snRNP occupancy of the strong 5′ splice site by psoralen crosslinking or immunoprecipitation with anti-U1A antisera, we did not observe any differences between RNAs lacking or containing any of the silencer elements (data not shown). While these approaches are useful for measuring the extent of U1 snRNP occupancy, they are not informative regarding potential differences in the manner in which U1 snRNP engages the affected 5′ splice site. To address this issue, we examined 5′ splice site recognition in the presence or absence of silencing elements using a nuclease protection assay of RNAs containing a single labeled phosphate 5′ of the uridine in the Gp*U at the 5′ splice site (Figure 5A and Maroney et al., 2000b). We have shown previously (Maroney et al., 2000a) that the binding of U1 snRNP to the strong 5′ splice site in the absence of any known splicing control element produces a characteristic pattern of nuclease resistant fragments (Figure 5B, none). Essentially identical patterns of protection were observed in the presence of silencing elements C and F (Figure 5B, Motifs C, F) indicating that these motifs did not markedly perturb U1 snRNP binding, at least as judged by this assay. However, strikingly distinct patterns of protection were observed in the presence of silencing elements A, B, D and E. All of the protections in the control RNA and those present in RNAs containing silencers B, C, D and F were strictly dependent upon U1 snRNP binding, since no protected fragments were observed when the 5′ end of U1 snRNA was blocked with a complementary 2′ Ome oligonucleotide. Extensive analyses, including RNA affinity approaches, have failed to reveal any candidate proteins that bind to these silencing elements (data not shown); observations which suggested the possibility that the U1 snRNP/5′ splice site interaction itself might be altered. Figure 5C provides direct evidence that this is the case for silencing element D; the distinctive pattern of protected fragments observed in nuclear extract was closely recapitulated using highly purified U1 snRNP (Hochleitner et al. 2005) (Figure 5C). The fact that the control RNA as well as two of the silencing elements (Motifs C and F) yield “wild type” patterns of protection clearly indicates that not all sequences elicit distinct nuclease resistant species and highlights the functional significance of those that do (see Discussion).
Intronic silencer A and exonic silencer E also showed altered patterns in the nuclease protection assay (Figure 6B, Motifs A and E, lanes +−), but with these RNAs we observed accumulations of U1 snRNP-independent fragments (Figure 6B, Motifs A and E, lane ++). Both of these silencers contain UAG motifs, characteristic of potential hnRNP A1 binding sites (e.g. Burd and Dreyfuss, 1994) and RNA affinity purification indicated that RNAs containing either silencer bound hnRNP A1 (Figure S5A and data not shown). Further analyses were conducted with exonic silencer Motif E. Figure 7A and 7B show that a GST-hnRNP A1 fusion protein (Blanchette and Chabot, 1999) specifically pulled down U1 snRNP only when Motif E was present and only when U1 snRNP was allowed to bind to the 5′ splice site. Because the fusion protein specifically crosslinks to an RNA containing element E (Figure 7C) it was possible to perform the reciprocal pulldown. Figure 7D shows that the crosslinked protein was immunoprecipitated by anti-U1A antisera but not by control antisera. These results provide direct evidence that U1 snRNP and hnRNP A1 bind to the same RNA molecules.
Here, we have employed a functional SELEX strategy in vitro to identify intronic and exonic sequence elements that silence nearby 5′ splice sites. Importantly, the silencers we characterized function both in vitro and in vivo. The data support several conclusions. First, effective silencing can be achieved without occluding or inactivating the affected site. Second, two of the six motifs (B and D) recovered from the selection may function idiosyncratically since they are not enriched in or near pseudoexons compared to real exons, yet still act as silencers in the construct used for selection. This result suggests the possibility that many splicing control elements may be relevant only in their specific pre-mRNAs and thus invisible to informatic techniques which rely on statistical analyses of motif frequencies. Third, informatic analyses indicate that four of the identified motifs (A, C, E and F) play a role in suppressing the use of both pseudo 5′ splice sites and pseudoexons (see Supplementary Material). Fourth, the results strongly suggest that intrinsic features of RNA sequences (e.g. motif D and perhaps B, C and F) themselves (without the participation of ancillary proteins) can influence the ability of 5′ splice sites to compete with each other. Fifth, we demonstrate that silencers can alter the U1 snRNP/5′ splice site complex in a manner that renders the affected site to be at a competitive disadvantage with respect to 5′ splice sites occupied by an “unaltered” U1 snRNP/5′ splice site complex (Figure 7). Sixth, and perhaps most importantly, we show that effects on non-rate limiting kinetic steps in splicing can lead to dramatic shifts in splice site choice.
In our experimental design, we focused on one construct under one set of conditions in extracts prepared from one cell type and analyzed only those motifs that gave the strongest phenotypes. It is important to note that we ignored sequences that produced substantive but less dramatic splice site switching. While we do not know how many sequence motifs would yield intermediate phenotypes, it seems likely that it is much larger than six. Furthermore, only one intronic and one exonic position were analyzed. It will be of interest to determine which sequences would emerge if the selections were carried out with the randomized sequence inserted at different positions or in different contexts. It will also be of interest to determine if sequences that could overcome the silencing effects could be selected in silencer containing constructs. Despite the inherent limitations of our current study, the results suggest that the regulation of splicing may be remarkably subtle and the “splicing code” (e.g. Black, 2003; Fu, 2004; Hertel, 2008; Matlin et al., 2005) correspondingly complex.
Intuitively, it would be expected that the most effective splicing silencers would sequester or otherwise inactivate the affected splice site. The silencers we have selected clearly do not function in this manner since the affected site remains fully functional but is not used when a competing site is present. When interpreted from a kinetic perspective, our observations rule out any scenario in which the silencers affect a slow or rate limiting step. If this were the case, presence of the silencers would cause reductions in the ability of the affected site to function proportional to the extent of activation of the competing site.
To account for the observed data, the silencers must affect a fast kinetic step to allow action of the silencer without substantive effects on the overall reaction rate. The fast reaction step has to be more than one order of magnitude faster than the slow reaction step when slowed by the silencer, and, consequently, several orders of magnitude faster than the rate limiting step when unaffected by the silencer. Introduction of the silencers at both splice sites would slow the fast reaction step at both sites and re-establish the original pattern of splice site selection, as we have observed (see Figures 3 and and77).
We suggest that such a fast step could be the joining of 5′ and 3′ splice sites in a complex committed to splice site choice. This interpretation is consistent with the findings of Lim and Hertel (2004) who showed that commitment to the general splicing pathway and commitment to specific splice site pairing are kinetically separable. Specifically, they demonstrated that commitment to splice site pairing became irreversible only upon formation of A complex, when U2 snRNP is locked onto the pre-mRNA. We suggest that the silencers function by altering the way in which U1 snRNP engages the 5′ splice site in such a way that it is less able to efficiently engage U2 snRNP, and thus the unaffected site obtains a competitive advantage. This interpretation is consistent with our nuclease protection data (see Figure 5) and with recent studies using site-directed hydroxyl-radical footprinting (Dönmez et al., 2007) which demonstrated that U1 and U2 snRNPs are in close proximity in early spliceosomal complexes. These studies suggested that the two snRNPs engage in a direct spatially-fixed interaction wherein the 5′ end of U2 snRNP is located on a specific “side” of U1 snRNP. Any distortion or misorientation of U1 snRNP caused by a silencer could negatively affect its ability to establish proper contact with U2 snRNP. The notion that appropriate alignment of U1 and U2 snRNPs early in spliceosome assembly is crucial for commitment to splice site choice provides a possible explanation for the observation that some of the silencers affect “weak” 5′ splice sites more than “strong” ones. We suggest that U1 snRNP bound to “weak” sites, i.e. those with fewer base pairs, are already somewhat “misaligned” and thus more susceptible to further distortion by the silencers.
In addition, it seems likely that hnRNP A1 may exert its silencing function in a similar manner; i.e. by affecting the alignment of U1 and U2 snRNPs. There have been several distinct models proposed for the mechanism by which hnRNP A1 can silence (reviewed in Black, 2003; Black and Graveley, 2006; Matlin et al., 2005). While all of these mechanisms may be valid in certain contexts, our finding that the protein binds simultaneously with U1 snRNP suggests that it may directly interfere with the ability of U1 snRNP to interact correctly with U2 snRNP either by altering the conformation of the U1 snRNP particle or perhaps by shielding important surfaces on the snRNP necessary for establishing correct alignment with U2 snRNP.
In summary, we suspect that a large fraction of examples of alternative and regulated splicing events will be dictated by kinetic parameters similar to the ones we have described. In this regard, there are several examples where the function of splicing control elements is only evident in the presence of competing splice sites (e.g. Cheah et al., 2007; Lam et al., 2003; Reed and Maniatis, 1986) and additional poorly understood examples where sequence context determines the use of competing sites (e.g. Chen and Helfman, 1999; Khelil et al., 2008; Krawczak et al., 2007; Manabe, et al. 2007; Mayeda and Ohshima, 1988; Nelson and Green, 1988; O’Neill et al., 1998, Ule et al., 2006).
In any multi-intronic pre-mRNA, each internal 5′ splice site is in competition with multiple 5′ splice sites as is each 3′ splice site. Myriad cross exon and cross intron interactions ensure that correct sites are paired. However, it is not hard to imagine that minor perturbations of any of these interactions [e.g. by tissue specific or developmental stage specific changes in the levels of splicing factors (Karni et al., 2007; Olson et al., 2007; Park et al., 2004), rates of transcription and/or chromatin structure] (reviewed in House and Lynch, 2008; Kornblihtt, 2006; Maniatis and Reed, 2002) could alter a delicate kinetic balance and result in the very complex patterns of splice site choice that are observed. Such a view makes sense from an evolutionary perspective in that small advantageous changes in intronic or exonic sequence could rapidly lead to the expansion of proteomic diversity seen in higher eukaryotes.
The construction of pre-mRNA templates containing duplicated 5′ splice sites comprised of portions of the rat preprotachykinin and Drosophila doubesex gene is described in detail in Supplemental materials, as is the introduction of fully randomized regions either +11 to +22 or −18 to −7 relative to the proximal 5′ splice site. All in vitro splicing reactions were conducted with body labeled substrate for 60 minutes in Hela cell nuclear extract (Dignam et al., 1983) as described (Maroney et al., 2000a) except where indicated differently in the text or Figure Legends. For the functional SELEX studies, the entire randomized pools of templates were transcribed; full length transcripts were gel purified and bound to purified U1 snRNP in NET-2 buffer. After immunoprecipitation with polyclonal anti-U1A antibody (Kambach and Mattaj, 1992), pre-mRNAs were deproteinized and added to in vitro splicing reactions. For debranching, lariat-3′ exon intermediates were gel purified and incubated with Hela cell cytoplasmic S100 extract under splicing conditions for 30 min at 30°C as described (Ruskin and Green, 1985).
Full length DNA fragments were cloned into pcDNA3.1 (Invitrogen) under the control of the CMV promoter and transfected into Hela cells using Lipofectamine2000 (Invitrogen) according to the manufacture’s protocol. 24 hr after transfection, the extent and position of splicing was determined by semi-quantitative RT-PCR as described in Supplemental Materials.
RNA fragments spanning the region from −45 to + 45 relative to the proximal 5′ splice site containing each of the individual silencer motifs or lacking any silencer, were uniquely labeled at G*U of the 5′ splice site. Site-specific labeling, oligonucleotide Inhibition, and nuclease protection assays followed the procedures described by (Maroney et al., 2000a; Maroney et al., 2000b). For details, see the Supplemental Materials.
Incubations containing nuclear extract either with or without added GST-hnRNP A1 as indicated in the Figure Legends were mixed with pre-washed glutathione sepharose 4B (Amersham Biosceinces) in NET-2 for 1 hour at 4°C. The bound complexes were washed three times, eluted with reduced glutathione (50 mM glutathione in 50 mM Tris-HCl, pH 8.0) and subjected to further analysis.
For the analysis of cross-linking to site-specifically labeled RNAs, reactions were irradiated with 254 nm UV light (see Supplemental materials) and then selected with glutathione sepharose as described above. The eluant was digested with 2 μg of RNase A at 37°C for 30 min before proteins wre fractionated on 10% polyacrylamide SDS gels.
For immunoprecipitation of cross-linked proteins, incubations, treated as above with UV light, were mixed with polyclonal anti-U1A antibody (Kambach and Mattaj, 1992) or control antibody, pre-bound to protein A agarose beads in NET-2 buffer supplemented with RNasin (Promega) for 1 hour at 4°C. The bound complexes were washed three times, digested with RNase A and then selected on glutathione sepharose. Proteins were eluted from the beads at 95°C with SDS running buffer. Isolated proteins were then resolved via SDS-PAGE and visualized by autoradiography.
We thank J. Bruzik, B. Chabot and P. Grabowski for providing constructs; S. Gunderson and I. Mattai for anti-U1A antiserum. We also thank D. Black, C. Burge, M. Caprara, X.D. Fu, B. Graveley, K. Hertel, B. Konforti, A. Krainer, K. Lynch, J. Manley, D. Rio, and J. Steitz as well as the anonymous reviewers for constructive comments, and A. M. Micenmacher for preparing the manuscript. The research was supported by DFG grants to R.L. and NIH grants to E.J., L.A.C. and T.W.N.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.