Functional crosstalk between exon enhancers, polypyrimidine tracts and branchpoint sequences

Abstract

We recently identified enhancer elements that activate the weak 3' splice site of alpha-tropomyosin exon 2 as well as a variety of heterologous weak 3' splice sites. To understand their mechanism of action, we devised an iterative selection strategy to identify functional pyrimidine tracts and branchpoint sequences in the presence or absence of enhancer elements. Surprisingly, we found that strong pyrimidine tracts were selected regardless of the presence of enhancer elements. However, the presence of enhancer elements resulted in the selection of multiple, non-consensus branchpoint sequences. Thus, enhancer elements apparently activate weak 3' splice sites primarily by increasing the efficiency of splicing of introns containing branchpoint sequences with less than optimal U2-branchpoint pairing arrangements. Comparison of consensus sequences from both our selection strategy and compilations of published intron sequences suggests that exon enhancer elements could be widespread and play an important role in the selection of 3' splice sites.

Characterization of human RNA splice signals by iterative functional selection of splice sites

“…After six rounds of selection the randomized region at positions 7-16 downstream of the BP sequence was significantly enriched for uridines (54%) and to a lesser extent cytosines (39%)+ Guanosines and adenosines occurred at 4% and 3% of the positions, respectively+ A similar, but slightly less pronounced pyrimidine enrichment was observed in the study by Buvoli et al+ (1997) over two rounds of selection+ At the 39 end of the randomized insert the pyrimidine rich region was followed by a single highly variable nucleotide and an invariant YAG motif+ Uridine residues are also more common than cytidine residues in pyrimidine tracts of natural introns (Shapiro & Senapathy, 1987) and the uridine content is an important determinant of the ability of the pyrimidine tract to compete for BP selection in a construct containing two polypyrimidine tracts adjacent to identical BP sequences (Coolidge et al+, 1997)+ However, our result indicates that the presence of some cytosines is stimulatory to splicing+ A search for common sequence patterns revealed a high frequency of UCCU motifs+ This sequence strongly resembles the UC 2-3 U motif selected in a U2AF 65 affinity selection experiment (Singh et al+, 1995), suggesting that U2AF 65 binding is an important factor in selection of competing splice sites+…”

“…In one selection experiment we selected BP sequences from a randomized 15-mer region that were able to support pre-mRNA splicing+ The randomized insert evolved towards a UACUAAC motif during seven rounds of selection, implying that pre-mRNAs with a BP sequence, optimized for base pairing with U2 snRNA, are favored in splicing+ Comparison of the sequence consensus in the individual pools revealed characteristic differences in selection rates of individual positions+ After only one round of selection the conservation was limited to UAA, suggesting that this sequence represents a minimal core branch site+ To go through one round of splicing the BP sequence must participate in base pairing with U2 snRNA and be able to expose a suitable nucleophile for the first transesterification step+ Some of the isolated and experimentally confirmed BP sequences after the first round of selection are only able to form 2-3 regular base pairs with U2 snRNA+ This interaction probably reflects a minimal requirement for base pairing for transesterification to occur+ Interestingly, the adenosine preceding the BP nucleotide is the least important in the UAA motif (See Fig+ 3A,E,F,G), suggesting that some structural flexibility is allowed at this position+ This may also explain the high frequency of branching to position Ϫ1 (relative to the BP) that we and others have observed (Hornig et al+, 1986;Query et al+, 1994;Buvoli et al+, 1997; this report)+ The importance of the BP adenosine and the uridine at the Ϫ2 position bears strong resemblance to the specificity observed for the essential splicing factor SF1 (Berglund et al+, 1997) that is required for the association of U2 snRNP with the pre-mRNA (Kramer, 1992)+ Thus, our data supports a model in which the interaction of SF1 and the base pairing of U2 snRNA with the BP sequence cooperate in the assembly of an active spliceosome+ Early recognition of the BP sequence by SF1 is supported by studies in yeast where it was shown that BBP (the yeast homolog of SF1) recognizes the BP sequence in complex E before entry of U2 snRNA (Berglund et al+, 1997;Pascolo & Séraphin, 1997)+ To be selected during the successive rounds of splicing, the sequence must be able to compete efficiently with the other RNA species in the pool for binding of splicing factors+ Our result suggests that an interaction between U2 snRNA and the pre-mRNA involving up to six Watson-Crick base pairs is an important determinant of the competitive strength of a given BP sequence+…”

“…In the selection of sequences surrounding the BP sequence only sequences from lariat intermediate were amplified during the amplification step+ Interestingly, all of the selected downstream sequences exhibited a conserved AG dinucleotide at the 39 end after six rounds of selection, suggesting a preference for an optimal 39 splice site element, even in the absence of specific second-step selection+ To examine the effect of the AG dinucleotide, the splicing of a transcript from clone 132 ( Fig+ 10B, lanes 2-4), whereas the AGto-AC mutation resulted in a decreased and delayed formation of complex A and an almost complete block of complex B formation (Fig+ 10B, lanes 7-8)+ The nonspecific complex H assembled equally well on both pre-mRNAs+ Thus, changing the conserved AG dinucleotide into AC severely affects splicing already before or at the time of U2 snRNP association+ DISCUSSION A comparative analysis of naturally occurring RNA splicing motifs is a powerful approach to identify functionally important nucleotides+ However, because splicing has evolved in vivo in the context of different premRNAs and under selection pressure from a complex set of functional requirements, more subtle sequence and structure preferences for splicing may not be detectable from alignments of natural sequences+ To obtain a simpler view of functional cis-acting sequences in RNA we have used a functional splicing/selection protocol and applied it to RNA libraries containing randomized sequences at selected positions+ A similar approach was used by Buvoli et al+ (1997) to study the effect of exon enhancer elements on branch site and polypyrimidine sequence selection+ However, in the present study a larger sequence space and number of selection rounds were employed, allowing the detection of more subtle sequence preferences+ A total of 47 nucleotide positions encompassing the BP region and 39 splice site were investigated+…”

Influence of Polymerase II Processivity on Alternative Splicing Depends on Splice Site Strength

“…Splicing enhancers can compensate for either a weak PPT or a weak branch point (23). When regulation of the splice sites selection is required, cells can modulate the production of different mRNAs by differentially expressing SR proteins that bind to splicing enhancer elements.…”

Dichotomous splicing signals in exon flanks

“…Iterative functional selections have not yet been applied toward a functional definition of ISE or ISS elements, although they have been used to define the binding sites of specific splicing inhibitory factors, e.g., PSI, PTB, and hnRNP A1 (Burd and Dreyfuss 1994;Singh et al 1995;Amarasinghe et al 2001). Iterative selections have been used to define functional polypyrimidine tracts and BP sequences (Buvoli et al 1997;Lund et al 2000). Although these are intronic sequences necessary for splicing, they are considered elements integral to the splicing process rather than enhancers.…”