the organization of 3' splice-site sequences in mammalian...

The organization of 3' splice-site sequences in mammalian introns

Robin Reed

Department of Cellular and Molecular Physiology, Harvard Medical School, Boston, Massachusetts 02115 USA

A model pre-mRNA substrate was used to carry out a detailed investigation of the functional organization of sequences at the 3' end of mammalian introns. This analysis revealed a difference in the sequence requirements for the first and second steps of the splicing reaction (lariat formation and exon ligation, respectively). Maximal efficiencies of lariat formation require a pyrimidine stretch directly adjacent to the branch site. In addition, efficient lariat formation can be specified in at least two distinct ways, one that requires the AG dinucleotide at the 3' splice junction, and the other that does not: If the pyrimidine stretch is short (14 nucleotides), an adjacent AG is essential; in contrast, the AG is not required in the presence of a long pyrimidine stretch (26 nucleotides). In a pre-mRNA containing a long pyrimidine stretch, efficient lariat formation is observed when the branch site is located greater than 100 nucleotides upstream from the AG or when the AG is preceded by a purine stretch. Although splicing usually takes place at the first AG downstream from the branch site, both distance and sequence play roles in the efficiency of this reaction.

[Key Words: 3' Splice site; pre-mgNA; branchpoint sequences; lariat formation]

Received August 22, 1989; revised version accepted October 24, 1989.

The sequences required for splicing in higher eukaryotes include conserved elements at the 5' and 3' splice sites and a less conserved element, the branchpoint sequence (BPS), at the site of lariat formation (for reviews, see Green 1986; Padgett et al. 1986; Maniatis and Reed 1987; Krainer and Maniatis, 1988; Steitz et al. 1988). In addition, exon sequences play a role in splice-site selection (Somasekhar and Mertz 1985; Reed and Maniatis 1986; Helfman et al. 1988; Hampson et al. 1989; Streuli and Saito 1989). Each of these sequence elements appears to have multiple roles in the splicing reaction. For example, sequences at the 5' splice site are required for cleavage at the 5' splice site and lariat formation, as well as for cleavage at the 3' splice site and exon ligation. Similarly, sequences at the BPS and 3' splice site are required for both of these steps (for review, see Krainer and Maniatis 1988; Steitz et al. 1988).

Conserved sequences at the 3' splice site consist of a 12-nucleotide pyrimidine stretch, followed by an AG dinucleotide, and a pyrimidine residue immediately preceding the AG (Mount 1982). The BPS is usually located 18-40 nucleotides upstream from the 3' splice junction (Ruskin et al. 1984; Zeitlin and Efstradiatis 1984; Keller and Noon 1984; Reed and Maniatis 1985). U2 small nuclear ribonucleoprotein particle (snRNP) binds to a region of the intron that includes the branch site (Black et al. 1985), and mutations in the BPS that decrease com- plementarity with a conserved region in U2 snRNA decrease the efficiency of splicing (Reed and Maniatis 1988; Zhuang et al. 1989). Recently, direct evidence was obtained that base-pairing between U2 snRNA and the

BPS occurs in higher eukaryotes (Wu and Manley 1989; Zhuang and Weiner 1989), as was shown previously in yeast (Parker et al. 1987). Binding of U2 snRNP to the branch site in metazoan introns is facilitated by a factor, designated U2AF, that binds to the pyrimidine stretch (Ruskin et al. 1988).

Several mutations in the sequence element at the 3' splice site have been examined during splicing in vitro (for reviews, see Green 1986; Padgett et al. 1986; Steitz et al. 1988; Krainer and Maniatis 1988); however, the role of these sequences in the first and second steps of the reaction has not been clearly established. Mutations in the AG dinucleotide in human f~-globin in t ron 1 decrease the efficiency of lariat formation, but do not abolish this reaction (Reed and Maniatis 1985; Ruskin and Green 1985). In contrast, deletion of the pyrimidine stretch in human ~-globin intron 1 or an adenovirus intron abolishes lariat formation (Frendewey and Keller 1985; Reed and Maniatis 1985; Ruskin and Green 1985) and spliceosome assembly (Frendewey and Keller 1985; Bindereif and Green 1986). These data suggested that the pyrimidine tract is essential for lariat formation and that the AG dinucleotide increases the efficiency of this reaction. Examination of mutations in the 3' splice site in other introns, however, has made it difficult to gener- alize these rules. In contrast to human f~-globin intron 1, mutations in the AG dinucleotide of rabbit ~3-globin intron 2 virtually abolish the first step of the reaction (Aebi et al. 1986) and spliceosome assembly (Lamond et al. 1987). Similarly, mutation of CAG to AAG in rabbit ~-globin intron 2 leads to a substantial reduction in the

GENES & DEVELOPMENT 3:2113-2123 © 1989 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/89 $1.00 2113

Cold Spring Harbor Laboratory Press on March 10, 2021 - Published by genesdev.cshlp.orgDownloaded from

http://genesdev.cshlp.org/

http://www.cshlpress.com

Reed

efficiency of the first step of the reaction (Aebi et al. 1986) whereas the corresponding mutation has no effect in human [3-globin intron 1 (R. Reed, unpubl.).

In contrast to these observations, the consensus 3' splice site does not appear to be required at all for the first step of the splicing reaction with some pre-mRNAs. For example, the splicing efficiency of an SV40 early region pre-mRNA is not significantly affected by insertion of a purine stretch adjacent to the AG dinucleotide (Fu et al. 1988). In addition, the 3' splice site of several naturally occuring pre-mRNAs consists of purine-rich sequences preceding the AG dinucleotide (Mount 1982; Falkenthal et al. 1985; Shelley et al. 1985; Bernstein et al. 1986; Rozek and Davidson 1986; George et al. 1989).

The functional significance of the conserved distance between the branch site and the 3' splice junction is also not clear. Most branch sites are located between 18 and 40 nucleotides from the 3' splice junction suggesting that branch-site selection occurs by a mechanism that measures the distance between the 3' splice junction and the BPS (for review, see Green 1986). However, this model is difficult to reconcile with the observation that BPSs located as far as 177 nucleotides upstream from the 3' splice junction are used efficiently in several alternatively spliced pre-mRNAs (Helfman and Ricci 1989; Smith and Nadal-Ginard 1989; E. Brody, pers. comm.; I. Mims and P. Bingham, pets. comm.). Additional studies are required to determine whether these alternatively spliced pre-mRNAs contain unique sequences or struc- tures for the use of distant branch sites, or whether the conserved distance between the AG and branch site seen in most pre-mRNAs is not a strict requirement for lariat formation.

Although a systematic analysis of the sequence requirements for cleavage at the 3' splice site and exon ligation has not been reported, in vitro splicing studies have shown that the AG dinucleotide is essential for this step of the reaction (Reed and Maniatis 1985; Ruskin and Green 1985; Aebi et al. 1986). In addition, analysis of a [3-thalassemia splicing mutant containing two AG dinucleotides revealed that the first AG downstream from the branch site is used (Krainer et al. 1985; Reed and Maniatis 1985). This observation suggests that proximity of the AG to the branch site is an important parameter in selecting the correct AG.

To investigate the role of both sequence and distance between the branch site and 3' splice junction, I examined the effects of deletions, duplications, and substitutions on the first and second steps of the splicing reaction. These studies revealed that distant branch sites, analogous to those seen in some alternatively spliced pre-mRNAs (Helfman and Ricci 1989; Smith and Nadal- Ginard 1989), can be activated in a model pre-mRNA substrate if a long pyrimidine stretch is located immediately downstream from the branch site. Efficient lariat formation was also observed in pre-mRNAs containing purine-rich 3' splice sites if a long pyrimidine stretch follows the branch site. However, when a short pyrimidine stretch follows the branch site, an adjacent AG dinucleotide is required. Finally, the data reveal that there

is a high degree of flexibility in the arrangement of 3' splice site sequences that promote efficient splicing.

Results

To investigate the sequence requirements for the first and second steps of the splicing reaction, pre-mRNAs containing mutations between the 3' splice junction and branch site were examined during splicing in vitro. To facilitate this analysis, a transcription template encoding a model pre-mRNA substrate was constructed (see Materials and methods). This pre-mRNA is derived from human [3-globin pre-mRNA and contains synthetic sequences extending from the branch site to the 3' splice junction of intron 1. The synthetic sequence contains a single BPS followed by a 28-nucleotide pyrimidine-rich sequence and AG dinucleotide (Fig. 1, pre-mRNA A). The branch-site adenine is located 32 nucleotides upstream of the 3' splice junction. The rationale for substi- tuting synthetic sequences for the normal sequences in this region was based on a number of considerations. First, adenine residues that could serve as potential cryptic branch sites were eliminated. Second, a portion of exon 2 that contains cryptic 3' splice sites was eliminated on the basis of previous studies showing that these sites are used efficiently when the normal 3' splice site is mutated (Krainer et al. 1985). Third, the synthetic sequence includes restriction sites for introducing mutated sequences. Examination of the in vitro splicing products generated from the model pre-mRNA demon- strated that this RNA precursor is spliced efficiently tFig. 1, lanes 1 and 2).

Purine substi tut ions adjacent to the BPS or A G dinucleotide decrease splicing efficiency

To study the role of the pyrimidine stretch, purine-rich sequences were substituted either adjacent to the BPS or adjacent to the AG dinucleotide. (Adenine nucleotides were not used for substitutions adjacent to the BPS because of the potential for creating cryptic branch sites.) When four pyrimidine residues near the BPS were substituted with guanine residues, a dramatic decrease in the first step of the splicing reaction was observed (Fig. 1, lanes 3 and 4). The second step of the reaction was not blocked by this substitution as the lariat intermediate was efficiently converted to spliced product by the later time point (Fig. 1, lane 4). When an additional two pyrimidine residues near the BPS were substituted with guanines, the efficiency of the first step of the splicing reaction was decreased even further, and, again, the second step was not blocked (Fig. 1, lanes 5 and 6). A substantial decrease in the first step of the reaction was also observed with a pre-mRNA in which most of the pyrimidines adjacent to the AG dinucleotide were substituted with purines (Fig. 1, lanes 7 and 8). Interest- ingly, the second step of the reaction was not prevented by this purine substitution. Thus, the pyrimidine tract adjacent to the 3' splice junction appears to affect the

2114 GENES & DEVELOPMENT




3' Splice-site sequences in mammalian introns

A B C D

60' 105' 60' 105' 60' 105' 60' 105' I II I [ - - 1 1 I

IZ3 EEl

C:3_ IZ3

1 2 3 4 5 6 7 8

-32 G U B PS AG

UCUCUUCCUUUgUCCUgUUCCCaUUUCC [ii~~~'~" ~'~':~"

-32 G U B PS AG

B ucUcugcguggguCCuguuCccauuucc ~ i i ~ i . . : ~ ~ " ~ ~-~i " ~

-32 G U BPS AG

C UgUCggCgUgggUCCUgUUCCCaUUUCC ~ . ~ ~ : : : ~ . ,~ .

-32 G U BPS

D UCUCUUCCUUUgUCCUggaaaUgggaaC AG

Figure 1. The effect of purine substitutions in the pyrimidine-rich sequences between the branch site and 3' splice junction. Splicing reactions containing 32p-labeled SP6 pre-mRNAs were carried out for 60 or 105 min, as indicated, and analyzed by electrophoresis on a 6.5% denaturing polyacrylamide gel. The structure of the splicing intermediates and products corresponding to each band is shown. The band marked by the arrow is an exonucleolytic degradation product that results from protection of the pre-mRNA at the branchpoint region (Krainer et al. 1984; Ruskin et al. 1984; Noble et al. 1989). The schematic diagrams show the sequence of the region between the branch site and the 3' splice site and the structure of each pre-mRNA. The pyrimidines are shown in uppercase letters, and the purines are shown in lowercase letters. (GU) 5' splice site; (BPS) branchpoint sequence (CUGAC); (AG) 3' splice junction. The number above BPS indicates the distance of the branch site from the AG dinucleotide. The boxes indicate human B-globin exon sequences; the line indicates the intron. Exon 1 is 155 nucleotides, exon 2 is 180 nucleotides, and the intron is 130 nucleotides. The pre-mRNAs that were spliced are as follows: A (Lanes 1 and 2); B {lanes 3 and 4); C {lanes 5 and 6); D (lanes 7 and 8).

efficiency of the first, but not the second, step of the splicing reaction.

The decreased efficiencies of lariat formation observed with the purine substitution mutants may indicate that the pyrimidine stretch must be contiguous with both the AG dinucleotide and BPS. Alternatively, a minimal length of pyrimidines may be required for the first step of the reaction, and the purine substitutions merely shorten the pyrimidines below this minimal length. To examine this question, the purine substitutions were analyzed in the presence of longer pyrimidine tracts. In these pre-mRNAs, the branch site is located 44 nucleotides from the 3' splice junction (Fig. 2, lanes 3-8). For comparison, the splicing products obtained with the original model pre-mRNA substrate are shown in Figure 2, lanes 1 and 2. Examination of the pre-mRNA containing six guanine substitutions adjacent to the BPS and a long pyrimidine stretch adjacent to the AG (Fig. 2, lanes 5 and 6) revealed a substantial increase in splicing efficiency relative to the same substitutions in the presence of the shorter pyrimidine stretch (Fig. 1, lanes 5 and 6). A similar increase in efficiency was observed when a long pyrimidine stretch was inserted into the pre-mRNA containing the purine-rich stretch adjacent to the AG

(cf. Figs. 1 and 2, lanes 7 and 8). Thus, the presence of a longer pyrimidine tract does indeed increase the splicing efficiency of a pre-mRNA containing purine substitutions. However, the splicing efficiency of a pre-mRNA containing a continuous pyrimidine stretch between the BPS and AG (Fig. 2, lanes 3 and 4) is significantly more efficient than the purine substitution adjacent to the BPS (Fig. 2, lanes 5 and 6) and slightly more efficient than the purine substitution adjacent to the AG dinucleotide (Fig. 2, lanes 7 and 8).

These data suggest that the location of the pyrimidine stretch with respect to the BPS and AG, as well as the length of the pyrimidine stretch, affect the first step of the splicing reaction. The reaction is efficient when a long pyrimidine tract is directly adjacent to the BPS. This efficiency is not significantly increased when the pyrimidine stretch is also directly adjacent to the AG (cf. Fig. 2, lanes 3, 4, 7, and 8). In contrast, the efficiency of the first step of the reaction is reduced when the pyrimidine stretch is separated from the BPS by purine-rich sequences. In this case, the first step of the reaction is significantly more efficient if a long, rather than a short, pyrimidine stretch is located downstream (cf. Figs. 1 and 2, lanes 5 and 6).

GENES & DEVELOPMENT 2115




Reed

A B C D :

60' 105 ' 60' 105! 60 ~ 1 0 5 ' 60' 105'

1 2 3 4 5 6 7 8

-32

G U BPS AG

A t iiii!iiiiiiiiiiii~iii~!i~i~i~iiii} UCUCUUCCUUUgUCCUgUUCCCaUUUCC {i!iiiiiiiiiiiiiiiiiii~iiii!i!ili!i~ili,i,;il

-44 G U BPS AG

B Iiiiiiiiiiiiii!i~ii!i!i!iiii!iiiiii} ucucuuccuuugucucuuccuuuguccuguucccauuucc liiiiiii!iiiiiiiiiiiii~iii!iiiii!i!i!iiiiiiii i ]

-44

G U BPS AG

c Ii!~i~!ii!i!iiiiiiiiiiii} ugocggogu~g~uouooocoouogoccugoocccaouoco li~iiii;iiiiiiiiiiii~iiiiili!iiiii~iiiiiiililill

-44

G U BP$ AG

o t~ i i i~ ! i i i i i i ! i i : : i i l ° ~ ° ° ° ° ° ° ° ° ° ~ ° ° ° ° ° ° ° ° ° ° u ° ° ° ° u ~ ° ~ ° Iiiiiiii~iiiiiiii~i~iiii!iii!i!ililt Figure 2. The effect of purine substitutions in the presence of a long pyrimidine tract. Splicing reactions containing 32p-labeled SP6 pre-mRNAs were carried out for 60 or 105 min, as indicated, and analyzed by electrophoresis on a 6.5% denaturing polyacrylamide gel. The structure of the splicing intermediates and products corresponding to each band is shown. The arrow indicates the exonucleolytic degradation product that results from protection of the pre-mRNA at the branchpoint region. The schematic diagrams show the sequence between the branch site and the 3' splice junction and the structure of each pre-mRNA. Designations are as described in the legend to Fig. 1. A {Lanes 1 and 2); B {lanes 3 and 4); C (lanes 5 and 6); D {lanes 7 and 8).

A long pyrimidine stretch is functionally equivalent to a short pyrimidine stretch followed by an AG

To examine the requirement for the AG in the first step of the reaction, pre-mRNAs containing either a short or long pyr imidine stretch adjacent to the BPS were tested in the presence or absence of an adjacent AG. As shown in Figure 3, the first step of the reaction is vir tually abolished when the BPS is followed by a 14-nucleotide py- r imidine stretch, but no AG (Fig. 3, lanes 1 and 2). After long exposures of the film, a very low level of the first step of the reaction can be detected (data not shown). Strikingly, however, efficient splicing was observed when the identical 14-nucleotide pyr imidine stretch is followed directly by an AG dinucleotide (Fig. 3, lanes 3 and 4; note that the sequences following the AG differ in the two pre-mRNAs but do not affect splicing efficiency).

In contrast to these results, the presence or absence of the AG dinucleotide has very little effect on the first step of the splicing reaction when the BPS is followed by a long pyr imidine stretch (Fig. 3, lanes 5 and 6). In fact, the efficiency of lariat formation is about the same as that observed wi th the control pre-mRNA (Fig. 3, lanes 7 and 8). Thus, for the first step of the reaction, a long pyrimidine stretch adjacent to the BPS is functionally equivalent to a short pyr imidine stretch followed by an AG dinucleotide. Interestingly, longer f i lm exposures of pre-mRNA D (data not shown) revealed a very low level of a linear RNA species that appears to result from use of an AG dinucleotide located wi th in the second exon (41 nucleotides from the branch site; Fig. 3). The inefficient use of this AG for the second step of the reaction may be attributed to the presence of a G residue preceding the AG, as GAG does not appear to function efficiently for the second step of the reaction, at least in some introns (Smith et al. 1989, see Discussion).






A B C D E

40' 70' 70' 70' 40' 70' 40' 70'

I II I r - - l l

c : : z ~ -

F-V-l-~ - -

1 2 3 4 5 6 7 8

-16 A CUGACUCUCUUCCUU UgUC~CGCUGG~GGUGAG

[E3 EEl

rT-I C3_

-18 CUGACU CUCU U CCU U Ug UCAGGG~,~UGGG~C~G&CGCIJGG~JGG~G~<G

-18 CUGA~UCUCUUCCUUUgUCAGUUCCC~UU~CCCUG~CGC~GUGGUG&G

-28 CUGACUCUCU UCCUU Ug UCCUUCUCUCUgUCGIACG~C~JG=G~G~G~JGAG

-32 CUGACUCUCUUCCU UUg UCCUg U UCCCaU U UCCAGG~CG~CUGG~JG(~UG&G

Figure 3. The role of the AG dinucleotide in the presence of a short or long pyrimidine stretch. SP6 pre-mRNAs were spliced in vitro for 40 or 70 min as indicated and fractionated by electrophoresis on a 6.5% denaturing polyacrylamide gel. The structure of the splicing intermediates and products corresponding to each band is shown. The sequences between the branch site and the 3' splice junction are shown below the autoradiograph. The distances between the branch site and the 3' splice junction (pre- mRNAs B, C, E) or second exon (pre-mRNAs A, D) are indicated. The sequence of a portion of the second exon is indicated by shadowed lettering. The BPS and AG are underlined. A (Lanes 1 and 2); B (lane 3); C (lane 4); D (lanes 5 and 6); E (lanes 7 and 8).

Activation of a branch site located far upstream from the 3' splice junction

In the experiments presented above, the effects of the purine subst i tut ions on the first step of the splicing reaction were examined. However, the direct effect of these mutat ions may be at the level of branch-site selection, on the basis of the observation that 3' splice-site sequences are required for interactions at the branch site (Black et al. 1985; Chabot and Steitz 1987; Kramer 1988; Ruskin et al. 1988). To investigate this possibility, pre- mRNAs containing tandemly duplicated BPSs adjacent to either purine- or pyrimidine-r ich sequences were examined (Fig. 4). Using this cis-competition assay, the effects of different sequences on branch-site selection are readily detected by a change in the pattern of branch-site use. The role of distance between the BPS and 3' splice junction in branch-site selection was also examined using the cis-competition assay as the duplicated BPSs were located as far as 116 nucleotides upstream from the 3' splice junction. Similar cis-competition assays were used previously to investigate the sequence requirements for splicing (Reed and Maniatis 1986, 1988; Zhuang et al. 1989).

Pre-mRNA D (Fig. 4, lanes 10-12) contains two BPSs, located 32 and 68 nucleotides upstream from the 3' splice junction. A long pyr imidine stretch is adjacent to the BPS located at -68 , whereas purine-rich sequences

are adjacent to the BPS located at - 3 2 . The AG dinucleotide is preceded by a short pyr imidine stretch (16 nucleotides). Interestingly, exclusive and efficient use of the BPS located at - 6 8 was observed wi th this pre- mRNA. The basis for this branch-site ass ignment is described in Materials and methods. Although the normal AG dinucleotide was used for the second step of the reaction, the kinetics were much slower, and the reaction was less efficient than that observed when the BPS is located 32 nucleotides from the 3' splice junction (Fig. 4, cf. lanes 1-3, and lanes 10-12). These data show that lariat formation can occur efficiently at distances > 1 8 - 4 0 nucleotides from the 3' splice junction. How- ever, the efficiency of the second step of the reaction is clearly affected by the increased distance and/or by the sequences between the branch site and the 3' splice junction.

These conclusions are supported further by the observation that a BPS located 116 nucleotides upstream from the 3' splice junction was used efficiently in pre-mRNA B, which contains four tandemly duplicated BPSs (Fig. 4, lanes 4 - 6 ; for explanation of branch-site assignment, see Materials and methods). The BPS located at - 116 is followed by a long pyr imidine stretch, whereas the downstream BPSs are all followed by purine-rich sequences. A 16-nucleotide pyr imidine stretch is immediately upstream from the AG. However, in contrast to the results obtained wi th the pre-mRNA D (Fig. 4, lanes





Reed

FFG, t l • m

r T ] ~ u • k m

A B

2 2.5 1 2 2.5 1 . . . . . I i : - i i

C D ¸

2 2 . s l 2 , 5 1

: ~il ~ • II =

E

2 2 ,5 hrs.

~ - -

C 2 _ - -

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-32 A CUGACUCUCUUCCUUUgUCCUgUUCCCaUUUCCAG~C

-116 CUGACUCUCUUCCUU UgUCCUUCUCUCUg UCgggCaCUGACUgUC, ggCgUggg UC, gggCaCUGACUgUCggC, g U

gggUCgggCaCUGACUgUC ggCgUgggUCCUgUUCCCaU UU CC~__GC~C

-116 CUGACUCUCUUCCUUUgUCCUUCUCUCUgUCgggCaCUGACUgUCggCgU gggUCgggCaCUGACUg UC-,ggC-,gU

gggUCgggCaCUGACUgUC ggCgUgggUCCUggaaaUgggaaCAGC~C

-68 CUGACUCUCUUCCUUUgUCCUUCUCUCUgUCgggCaCUGACUgUCggCgUgggUCCUgU UCCCaUUUCCAG~I~,O

-68 CUGACUCUCUUCCUUUgUCCUUCUCUCUgUCgggCaCUGACUgUCggCgUgggUCCUggaaaUgggaaCAGC=~C

Figure 4. The role of distance and sequence between the BPS and 3' splice junction on the two steps of the splicing reaction. SP6 pre-mRNAs were spliced in vitro for 1, 2, or 2.5 hr, as indicated, and fractionated by electrophoresis on a 6.5% denaturing polyacrylamide gel. The structure of the splicing intermediates and products corresponding to each band is shown. The pre-mRNA and lariat intermediate are indicated at left for pre-mRNA A (lanes 1-3) and at right for the remaining pre-mRNAs. (*) The exonucleolytic degradation product that results from protection of the pre-mRNA at the branchpoint region. A linear RNA species that decreases during the splicing reaction is seen above the branchpoint protection band with pre-mRNAs C and E (lanes 7-9 and 13-15). This RNA species may be a stable breakdown product of these similar pre-mRNA substrates. The sequences between the branch site and 3' splice site are shown below the autoradiograph. The duplicated BPSs are underlined, and the BPS that is used for lariat formation is indicated by underscoring and boldface lettering. The distance between this BPS and the AG is indicated. Exon sequences are designated by shadowed lettering. A (Lanes 1-3); B (lanes 4-6); C (lanes 7-9); D (lanes 10-12); E (lanes 13-15).

10-12), barely detectable levels of the second step of the reaction were observed wi th pre-mRNA B (Fig. 4, lanes 4-6). This result suggests that the distance between the BPS and the 3' splice junction, the sequence, or both decrease the efficiency of the second step of the reaction. To address these possibilities, a purine stretch was subst i tuted immedia te ly upstream of the 3' splice site in pre-mRNAs B and D to generate pre-mRNAs C and E (Fig. 4). Interestingly, although the first step of the reaction was not affected by these substitutions, the second step of the reaction was abolished (Fig. 4, lanes 7-9 , 13-15). These data indicate that the sequences between the branch site and the 3' splice junction can affect the second step of the reaction. However, distance also appears to play an important role in the efficiency of the

second step of the reaction because the same purine substi tutions have no effect on the second step when the branch site and the 3' splice site are closer together (cf. Figs. 1, 2, and 4).

Further insight into the requirements for branch-site selection was provided by an analysis of a pre-mRNA in which a short pyr imidine stretch was adjacent to the upstream BPS (located at -56), and purine-rich sequences were located adjacent to the downstream BPS (Fig. 5a, lanes 9 and 10). Only very low levels of spliced RNA were detected wi th this pre-mRNA. Comparison of the splicing reaction of this pre-mRNA on a 6.5% and 8% denaturing polyacrylamide gel identified two lariat intermediate species, corresponding to use of both the BPS at - 5 6 and the BPS at - 3 2 (Fig. 5a, b, lanes 9 and 10).

2 1 1 8 GENES & DEVELOPMENT





a

W

~ B

A B C D E

60' 105' 60' 105' 60' 105' 60' 105' 60' 105' l ii it it il i

A B C D E

=,~ 60' los' 60' 105' 60' 105' 60' los' 60' ~os' U

m "k ] c::zc~ -]rv, ,~

,qt

m c z ~

- - rv -1

1 2 3 4 5 6 7 8 9 10

-32 CUGACUCUCUUCCUUUgUCCUgUUCCCaUUUCCA.~GG~C

1 2 3 4 5 6 7 8 9 10

-32 C~GACUCUCUUCCUUUgUCUCg~-GGGgUUUCUggUgggCacUGA~UCUCUUC~uUUgU~CUgUUCCCaUUUCCAG~'~;

-32 CUGACUgUCggCgUgggUCUCg~.~gUUUCUggUgggCaCUGACUCUCUUCCUUUgUCCUgUUCCCaUUUCCAGG~C

-21 CUGACUCUCU UCCU U UgUCUCgA.GgU UUC UgcA_GggCaCUGACUgUCggC, gUgggUCCUg U UCCCaU UUCCAG~C

-56 - ~ CUGACUCUCUUCCUUUgUCgggCaCUGACUgUCggCgUgggUCCUgUUCCCaUUUCCAGG=AC;

Figure 5. The effect of adjacent sequences on the relative use of tandemly duplicated branch sites. SP6 pre-mRNAs were spliced in vitro for 60 or 105 min, as indicated, and fractionated by electrophoresis on a 6.5% (a) or an 8% (b) denaturing polyacrylamide gel. The structure of the pre-mRNA and splicing products corresponding to each band is shown. (*) Lariat-exon intermediates; the arrow refers to the exonucleolytic breakdown product generated by protection at the branch site. The sequences of the relevant regions are shown below the autoradiograph. The functional BPS and AG are indicated by underscoring and boldface lettering, and the distance between them is shown. The duplicated AG and BPSs that are not used for splicing are underlined and exon sequences are designated by shadowed lettering. A (Lanes 1 and 2); B (lanes 3 and 4); C (lanes 5 and 6); D (lanes 7 and 8); E (lanes 9 and 10).

These lariat intermediates were not observed on the 6.5% gel because they comigrated wi th the pre-mRNA or its breakdown products, but were readily detected on the 8% gel. These results provide further evidence that the pattern of branch-site selection can be dramatical ly altered by the sequences adjacent to the BPS. In addition, the data show that efficient use of a BPS located far upstream from the 3' splice junction requires an adjacent pyrimidine stretch of a m i n i m a l length.

Data presented above (see Fig. 3) suggest that the use of the - 56 BPS, adjacent to the short pyr imidine stretch (Fig. 5a, b, lanes 9 and 10) should be improved by insertion of an AG dinucleotide, because this should com- pensate for the short pyr imidine stretch. To test this possibility, pre-mRNA D was examined {Fig. 5a, b, lanes

7 and 8). Because of the restriction sites used for con- structing this pre-mRNA, two AG dinucleotides are present. Analysis of the mobi l i ty of the lariat intermediate generated from pre-mRNA D revealed that the upstream BPS was used exclusively and efficiently (com- pare the mobi l i ty of this lariat intermediate wi th the lariat intermediates in lanes 3 - 6 in which the downstream BPS is used; see below). In addition, both AG dinucleotides present in pre-mRNA D were used wi th about equal efficiencies for the second step of the splicing reaction. Thus, as predicted, the AG dinucleotide can dramatical ly affect branch-site selection in the presence of a short pyr imidine stretch.

In all of the pre-mRNAs containing duplicated BPSs described above, the downstream BPS, adjacent to





Reed

purine-rich sequences, was either not used or was used very inefficiently. To demonstrate that these effects re- sulted from the presence of the adjacent purine stretch, rather than the location of the BPS within the pre- mRNA, a pre-mRNA was examined in which the upstream BPS was adjacent to the purine-rich sequences and the downstream BPS was adjacent to pyrimidines. As shown in Figure 5 (a,b, lanes 5 and 6), the spliced RNA product generated from this pre-mRNA is the same size as the spliced RNA generated from the control pre-mRNA (lanes 1 and 2). In addition, one lariat-exon RNA species was observed with this pre-mRNA. These data indicate that the downstream BPS is used as the branch site in this lariat intermediate. Thus, the downstream BPS is capable of functioning when adjacent to pyrimidines, confirming the importance of a pyrimidine stretch directly adjacent to the BPS for efficient branch- site selection.

The downstream BPS was also used exclusively and efficiently in a pre-mRNA in which the upstream BPS is adjacent to a short pyrimidine stretch and AG dinucleotide, whereas the downstream BPS is located next to a long pyrimidine stretch and an AG dinucleotide (Fig. 5a, b, lanes 3 and 4). The observation that the downstream site is used exclusively with this pre-mRNA suggests that a longer pyrimidine stretch is significantly more efficient than a shorter pyrimidine stretch, even in the presence of the AG dinucleotide.

Discussion

Previous studies suggested that the functional compo- nents at the 3' end of most introns in higher eukaryotes include a branchpoint sequence located at a conserved distance from the 3' splice junction and a 12-nucleotide pyrimidine tract adjacent to the AG dinucleotide. How- ever, the data presented in this paper reveal that sequence arrangements deviating significantly from this canonical 3' splice site are efficiently recognized by the splicing machinery. In addition, different sequence requirements were identified for the first and second steps of the splicing reaction, suggesting that different factors interact with 3' splice site sequences during each step of the reaction. Finally, these studies show that the distance between the branch site and the 3' splice junction affects not only the first but also the second step of the splicing reaction, depending on the sequences in this region.

3' Splice site sequences required for efficient lariat formation

Two distinct 3' splice-site sequence arrangements were identified that promote efficient lariat formation. These include a short pyrimidine stretch (14 nucleotides), followed by an essential AG dinucleotide and a long pyrimidine stretch (26 nucleotides) without an AG requirement. In both cases, the pyrimidine stretch must be lo-

cated immediately downstream from the branch site for maximal efficiencies of lariat formation.

A number of other functional 3' splice-site sequence arrangements are essentially derivatives of the sequence arrangement in which a long pyrimidine stretch replaces the AG requirement. For example, distant branch sites, located 68 and 116 nucleotides from the 3' splice junction, are used efficiently when a long pyrimidine stretch follows the branch site. Interestingly, a very extensive pyrimidine stretch follows the distant branch site in the alternatively spliced a- and [3-tropomyosin pre-mRNAs (Helfman and Ricci 1989; Smith and Nadal-Ginard 1989), and this pyrimidine stretch, rather than the 3' splice site, is required for lariat formation (Smith et al. 1989; D. Helfman, pers. comm.). Similar long pyrimidine stretches are also found adjacent to the distant branch sites in other alternatively spliced pre-mRNAs (E. Brody, pers. comm.; I. Mims and P. Bingham, pers. comm.). Thus, one mechanism for the use of distant branch sites may require the presence of a long pyrimidine stretch adjacent to the branch site.

The presence of a long pyrimidine stretch adjacent to the branch site also results in efficient lariat formation in introns containing purine-rich 3' splice sites. This sequence arrangement may explain the efficient use of some of the purine-rich 3' splice sites in naturally occur- ring pre-mRNAs (Mount 1982; Falkenthal et al. 1989; Shelley et al. 1985; Bernstein et al. 1986; Rozek and Da- vidson 1986; George et al. 1989). In addition, a pyrimidine stretch following the branch site was required for efficient splicing when purine-rich sequences were inserted next to the AG in an SV40 intron (Fu et al. 1988).

The pattern of branch site use observed with the pre- mRNAs containing tandemly duplicated BPSs shows that the pyrimidine content of the sequences immediately adjacent to the BPS is a critical determinant of branch-site selection. These data may therefore explain why particular cryptic or normal BPSs are preferred over others. For example, the cryptic branch site (TAC- TAAA), rather than the normal branch site (TACTAAC), was used when the yeast actin intron was spliced in a HeLa cell nuclear extract (Ruskin et al. 1986). Examina- tion of this pre-mRNA shows that the cryptic BPS is adjacent to pyrimidine-rich sequences (TACAAACTTTT- TaTTTTgTaTTgCTTTT), whereas the normal BPS is followed by more purine-rich sequences (TACTAA- CaagTTgaaTTgCaTTTaCaaa). Similarly, the distant branch sites used in the alternatively spliced oL- and [3- tropomyosin pre-mRNAs (Helfman and Ricci 1989; Smith and Nadal-Ginard 1989) are adjacent to long, un- interrupted pyrimidine stretches, whereas the nonfunc- tional BPSs downstream are followed by less pyrimidine-rich sequences. In contrast to higher eukaryotes, a pyrimidine stretch does not appear to be required for branch-site selection in yeast (Rymond and Rosbash 1985; Rymond et al. 1987).

Although the length of the pyrimidine stretch was identified as an important parameter in determining the efficiently of lariat formation in the pre-mRNAs examined here, the quality of the pyrimidine stretch, whether





it is continuous or interrupted by purines, is also likely to affect the efficiency of this reaction. In addition, the specific length and sequence requirements for the pyrimidine stretch in different pre-mRNAs may be affected by other parameters, such as the presence of multiple branch sites or the match of the branchpoint sequence to the consensus.

3' Splice site sequences required for efficient exon ligation

The sequences at the 3' site can affect the efficiency of the second step of the reaction. The magnitude of this effect, however, depends on the distance between the branch site and the 3' splice junction. When the BPS is within 18 and 40 nucleotides of the 3' splice junction, exon ligation occurs efficiently whether purine- or pyrimidine-rich sequences are adjacent to the BPS or AG dinucleotide. When the distance between the branch site and AG is increased, however, the sequences in between have a dramatic effect on the second step of the reaction. For example, when the branch site is located 68 nucleotides upstream from the AG, efficient exon ligation is observed when a pyrimidine stretch precedes the AG; in contrast, the second step of the reaction is abolished when this pyrimidine stretch is replaced with purines. Indeed, maximal efficiencies of the second step of the reaction may require continuous pyrimidine-rich sequences between distant branch sites and the 3' splice junction. This possibility is suggested by the observation that pyrimidine-rich sequences span the region between the BPS and 3' splice junction in all of the alternatively spliced pre-mRNAs that use distant branch sites (Helfman and Ricci 1989; Smith and Nadal-Ginard 1989). Data presented here suggest that such an extensive pyrimidine stretch is not essential for the first step of the reaction but, instead, may be required for the second step. A decrease in the efficiency of the second step of the reaction was also observed when purines replaced the pyrimidine stretch in an SV40 pre-mRNA (Fu et al. 1988).

In most introns, a pyrimidine residue is located immediately upstream of the AG, adenine residues occur less frequently, and guanine residues are rare (Mount 1982). Although a C residue preceded the AG in most of the introns examined in this study, a GAG followed closely (8 nucleotides) by a CAG was present in one intron. Both of these AG dinucleotides appear to be used with ap- proximately equal efficiency. Previous studies suggested that the AG nearest to the branch site is used for the second step of the reaction (Reed and Maniatis 1985). The use of both AGs in the study reported here may occur because GAG does not function efficiently for the second step of the reaction, leading to splicing at the more efficient CAG downstream. Inefficient use of GAG was observed previously in studies of the ~-tropomyosin intron (Smith et al. 1989). Alternatively, multiple AGs may function for the second step of the reaction if the AGs are clustered. Similar use of closely spaced AGs was also observed when multiple AGs were inserted at


the 3' splice site of an SV40 (Fu et al. 1988) and adenovirus (Ulfendahl et al. 1989) intron. In these cases, the upstream AG was preceded by adenine and guanine residues, respectively.

Interactions between splicing factors and the 3' splice site

Several different factors appear to interact with sequences at the 3' splice site (Kramer 1988), including three different heteronuclear RNP (hnRNP) proteins (Swanson and Dreyfuss 1988), U1 snRNP (Zillman et al. 1987), and two different proteins, designated U2AF and intron-binding protein (IBP). U2AF is thought to mediate U2 snRNP binding at the branch site (Ruskin et al. 1988), whereas IBP may mediate U5 snRNP binding at the 3' splice site (Gerke et al. 1986; Tazi et al. 1986). Neither IBP nor U2AF binds to pre-mRNAs lacking the pyrimidine tract. Both of these factors also bind less efficiently to a pre-mRNA lacking the AG dinucleotide, although IBP binding is affected more severely by this mutation (Gerke et al. 1986; Tazi et al. 1986; Ruskin et al. 1988). Several observations suggest that U2 snRNP binding precedes binding of U5 snRNP during spliceosome assembly (Grabowski and Sharp 1986; Konarska and Sharp 1986, 1987; Bindereif and Green 1987; Kramer 1988). However, the precise functional relationships between these factors, as well as the roles of the other factors that bind the 3' splice site, have not been established.

The data presented in this study are consistent with a model for branch-site selection that involves coopera- tive binding of a splicing factor(s), such as U2AF, to sequences at the 3' splice site. According to this model, pyAG constitutes a strong binding site for this factor, whereas py alone is a weak binding site (py and pur indicate short pyrimidine and purine stretches, respectively). The sequence arrangements, BPSpypyAG and BPSpypy could therefore bind two factors cooperatively whereas only one factor could bind to BPSpyAG. Very inefficient binding would be expected with BPSpy because one weak site is present. The relative strengths of 3' splice site sequences derived from data presented in this study are BPSpypyAG > BPSpyAG = BPSpypy > BPSpy. These relative strengths are the same as those predicted by cooperativity. The additional observation provided by this study, that branch-site selection is more efficient when the pyrimidine stretch is directly adjacent to, rather than separated from, the BPS (e.g., BPSpyAG versus BPSpurpyAG) could be explained if the factors bound to the pyrimidine stretch require direct contact with a factor, such as U2 snRNP, bound at the branch site. The pyrimidine stretch also appears to be required at a minimal distance from the AG dinucleotide for efficient cleavage at the 3' splice site and exon ligation. This reaction may involve a similar mechanism re- quiring direct contact between factors bound to the pyrimidine stretch and factors bound to the AG dinucleotide.

Sequences between the branch site and 3' splice junc-





Reed

t ion are essent ia l for the early steps of sp l iceosome assembly. These sequences have been s h o w n to affect the choice of 5' splice si tes in a l t e rna t ive ly spliced pre- m R N A s (Fu et al. 1988) and to be involved in the regula- t ion of a l te rna t ive spl ic ing (Smith and Nada l -Ginard 1989). T h e h igh degree of var iab i l i ty in func t iona l 3 ' spl ice-si te sequence a r rangements suggests tha t th is region of the in t ron m a y also play a central role in correct spl ice-si te se lec t ion in p re -mRNAs con ta in ing mul t ip l e in t rons .

Materials and methods

Plasmid constructions

SP64 plasmids used in this study were constructed from a derivative of plasmid H[3ABam (Reed and Maniatis 1986), which contains an XhoI linker in the MboII site (located at + 77 relative to the 5' end of intron 1). Initially, a parental clone was constructed by inserting an oligonucleotide into the XhoI site and the HinfI site ( + 165 relative to the 5' end of the intron) of the Hf~ABam derivative. The sequence of this oligonucleotide is as follows: 5'-TCGAGGTTTCTGGTGGGCACTGACTCTCT- TCCTTTGTCGACGCTGGTGGTG-3'. To construct the plasmids used in Figures 1-5, appropriate oligonucleotides were inserted into a HincII site located within the oligonucleotide used to construct the parental plasmid. Refer to the figures for relevant sequence information. Further details on the plasmid constructions are available on request. The nucleotide sequence of the relevant regions of each plasmid was verified by chain-termination DNA sequencing (Sanger et al. 1977).

Pre-mRNA synthesis and in vitro splicing

SP64 plasmids were linearized with BamHI for in vitro transcription. Pre-mRNA synthesis using SP6 polymerase was per- formed as described (Melton et al. 1984). In vitro splicing reactions were carried out as described by Krainer et al. (1984), ex- cept that polyvinyl alcohol was omitted. The reactions contained 5 - i 0 ng pre-mRNA and were incubated for the time lengths indicated in each figure legend.

Branch-site assignments

The conclusion that the upstream BPS is used for pre-mRNAs D and E in Figure 4 was based on the observation that the mo- bility of the lariat intermediates of these pre-mRNAs is the same as the lariat intermediate of pre-mRNA D in Figure 5, in which the upstream BPS is used. Note that these three pre- mRNAs are the same size and that the duplicated BPSs are located in the same positions. The conclusion that the upstream BPS is used for pre-mRNAs B and C in Figure 4 is based on the size of the lariat intermediate and on an oligonucleotide-di- retted RNase H assay used previously to map the structure of lariat intermediates (Ruskin et al. 1984; data not shown).

Acknowledgments

I thank Tom Maniatis for useful discussions and comments on the manuscript and Adina Chelouche for excellent technical assistance and help in preparation of the manuscript. HeLa cells used for nuclear extracts were provided by the National Insti- tutes of Health cell culture facility at MIT. I am grateful to Alex Nussbaum, Peter Gentile, and Gene Brown for oligonucleotide synthesis. R.R. is a Lucille P. Markey Scholar. This work was


supported by a grant from the Lucille P. Markey Charitable Trust.

Note added in proof

Using an in vivo cis competition assay in yeast, it has been shown that use of an alternative 3' splice site can be enhanced by increasing the uridine content preceding the AG dinucleotide; further, 3' splice sites preceded only by A-rich sequences are inhibited at the second step of splicing (B. Patterson and C. Guthrie, in prep.).

References

Aebi, M., H. Hornig, R.A. Padgett, J. Reiser, and C. Weissman. 1986. Sequence requirements for splicing of higher eukary- otic nuclear pre-mRNA. Cell 47: 555-556.

Bernstein, S.I., C.J. Hansen, K.D. Becker, D.R. Wassenberg II, E.S. Roche, J.J. Donady, and C.P. Emerson. 1986. Alternative RNA splicing generates transcripts encoding a thorax-specific isoform of Drosophila melanogaster myosin heavy chain. Mol. Cell. Biol. 6" 2511-2519.

Bindereif, A. and M.R. Green. 1986. Ribonucleoprotein complex formation during pre-mRNA splicing in vitro. Mol. Cell. Biol. 6: 2582-2592.

1987. An ordered pathway of snRNP binding during mammalian pre-mRNA splicing complex assembly. EMBO J. 6" 2415-2424.

Black, D.L., B. Chabot, and J.A. Steitz. 1985. U2 as well as U1 small nuclear ribonucleoproteins are involved in pre-messenger RNA splicing. Cell 42: 737- 750.

Chabot, B. and J.A. Steitz. 1987. Multiple interactions between the splicing substrate and small nuclear ribonucleoproteins in spliceosomes. Mol. Cell. Biol. 7" 281-293.

Falkenthal, S., V.P. Parker, and N. Davidson. 1985. Develop- mental variations in the splicing pattern of transcripts from the Drosophila gene encoding myosin alkali light chain result in different carboxyl-terminal amino acid sequences. Proc. Natl. Acad. Sci. 82" 449-453.

Frendewey, D. and W. Keller. 1985. The stepwise assembly of a pre-mRNA splicing complex requires U snRNPs and specific intron sequences. Cell 42: 355-367.

Fu, X.-Y., H. Ge, and J.L. Manley. 1988. The role of the polypyr- imidine stretch at the SV40 early pre-mRNA 3' splice site in alternative splicing. EMBO J. 7:809-817.

George, E.L., M.B. Ober, and C.P. Emerson. 1989. Functional domains of the Drosophila melanogaster muscle myosin heavy-chain gene are encoded by alternatively spliced exons. Mol. Cell. Biol. 9: 2957-2974.

Gerke, V. and J.A. Steitz. 1986. A protein associated with small nuclear ribonucleoprotein particles recognizes the 3' splice site of premessenger RNA. Cell 47" 973-984.

Grabowski, P.J. and P.A. Sharp. 1986. Affinity chromatography of splicing complexes: U2, US, and U4 + U6 small nuclear ribonucleoprotein particles in the spliceosome. Science 233: 1294-1299.

Green, M.R. 1986. Pre-mRNA splicing. Annu. Rev. Genet. 20: 671-708.

Hampson, R.K., L.L. Follette, and F.M. Rottman. 1989. Alterna- tive processing of bovine growth hormone mRNA is in- fluenced by downstream exon sequences. Mol. Cell. Biol. 9: 1604-1610.

Helfman, D.M. and W.M. Ricci. 1989. Branch point selection in alternative splicing of tropomyosin pre-mRNAs. Nucleic Acids Res. 17: 5633-5650.




Helfman, D.M., W.M. Ricci, and L.A. Finn. 1988. Alternative splicing of tropomyosin pre-mRNAs in vitro and in vivo. Genes Dev. 2:1627-1638.

Keller, E.B. and W.A. Noon. 1984. Intron splicing: A conserved internal signal in the introns of animal pre-mRNAs. Proc. Natl. Acad. Sci. 81: 7417-7420.

Konarska, M.M. and P.A. Sharp. 1986. Electrophoretic separa- tion of complexes involved in the splicing of precursors to mRNAs. Cell 46: 845-855.

1987. Interactions between small nuclear ribonucleoprotein particles in formation of spliceosomes. Cell 49: 763-774.

Krainer, A.R., T. Maniatis, B. Ruskin, and M.R. Green. 1984. Normal and mutant human f~-globin pre-mRNAs are faith- fully and efficiently spliced in vitro. Cell 36: 993-1005.

Krainer, A., R. Reed, and T. Maniatis. 1985. Mechanisms of human f~-globin pre-mRNA splicing. In The Robert A. Welch Foundation conferences on chemical research, ge- netic chemistry: The molecular basis of heredity, vol. XXIX, pp. 353-382. Houston.

Krainer, A. and T. Maniatis. 1988. RNA splicing. In Frontiers in transcription and splicing. (ed. B.D. Haines and D.M. Glover), pp. 131-206. IRL Press, Oxford and Washington D.C.

Kramer, A. 1988. Presplicing complex formation requires two proteins and U2 snRNP. Genes Dev. 2:1155-1167.

Lamond, A.I., M.M. Konarska, and P.A. Sharp. 1987. A muta- tional analysis of spliceosome assembly: Evidence for splice site collaboration during spliceosome formation. Genes Dev. 1: 532-543.

Maniatis, T. and R. Reed. 1987. The role of small nuclear ribonucleoprotein particles in pre-mRNA splicing. Nature 325: 673-678,

Melton, D.A., P.A. Krieg, M.R. Rebagliati, T. Maniatis, K. Zinn, and M.R. Green. 1984. Efficient in vitro synthesis of biologi- cally active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Res. 12: 7035-7056.

Mount, S.M. 1982. A catalogue of splice junction sequences• Nucleic Acids Res. 10: 459-472.

Noble, J.C.S., H. Ge, M. Chaudhuri, and J.L. Manley. 1989. Factor interactions with the SV40 early pre-mRNA influ- ence branchpoint selection and alternative splicing. Mol• Cell. Biol. 9: 2007-2017.

Padgett, R.A., P.J. Grabowski, M.M. Konarska, S. Seiler, and P.A. Sharp. 1986. Splicing of messenger RNA precursors. Annu• Rev. Biochem. 55: 1119-1150.

Parker, R., P.G. Siciliano, and C. Guthrie. 1987. Recognition of the TACTAAC box during mRNA splicing in yeast involves base pairing to the U2-1ike snRNA. Cell 49: 229-239.

Reed, R. and T. Maniatis. 1985. Intron sequences involved in lariat formation during pre-mRNA splicing. Cell 41: 95- 105.

• 1986. The role of exon sequences in splice site selection. Cell 46: 681-690.

• 1988. The role of the mammalian branchpoint sequence in pre-mRNA splicing. Genes Dev. 2: 1268-1276.

Rozek, C.E. and N. Davidson. 1986. Differential processing of RNA transcribed from the single-copy Drosophila myosin heavy chain gene produces four mRNAs that encode two polypeptides. Proc. Natl. Acad. Sci. 83: 2128-2132.

Ruskin, B., A. Krainer, T. Maniatis, and M.R. Green. 1984. Ex- cision of an intact intron as a novel lariat structure during pre-mRNA splicing in vitro. Cell 38: 317-331.

Ruskin, B. and M.R. Green. 1985. Role of the 3' splice site consensus sequence in mammalian pre-mRNA splicing. Nature


317: 732-734• Ruskin, B., C.W. Pikielny, M. Rosbash, and M.R. Green. 1986.

Alternative branchpoints are selected during splicing of a yeast pre-mRNA in mammalian and yeast extracts. Proc. Natl. Acad. Sci. 83: 2022-2026.

Ruskin, B., P.D. Zamore, and M.R. Green• 1988. A factor, U2AF, is required for U2 snRNP binding and splicing complex assembly. Cell 52: 207-219.

Rymond, B.C. and M. Rosbash. 1985. Cleavage of 5' splice site and lariat formation are independent of 3' splice site in yeast mRNA splicing. Nature 317: 735-737.

Rymond, B.C., D.D• Torrey, and M. Rosbash. 1987. A novel role for the 3' region of introns in pre-mRNA splicing of Saccha- romyces cerevisiae. Genes Dev. 1: 238-246.

Sanger, F., S. Nicklen, and A. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. 74: 5463-5467.

Shelley, C.S., C.R. Sharpe, F.E. Baralle, and C.C. Shoulders. 1985. Comparison of the human apolipoprotein genes. J. Mol. Biol. 186: 43-51.

Smith, C.W.J. and B. Nadal-Ginard. 1989. Mutually exclusive splicing of a-tropomyosin exons enforced by an unusual lariat branchpoint location: Implications for constitutive splicing. Cell 56: 749-758.

Smith, C.W.J., E.B. Parro, J.G. Patton, and B. Nadal-Ginard. 1989. The 3' splice site of mammalian introns is defined by scanning from an independently specified branchpoint. Na- ture (in press).

Somasekhar, M.B. and J.E. Mertz. 1985. Exon mutations that affect the choice of splice sites used in processing the SV40 late transcripts. Nucleic Acids Res. 13: 5591-5609.

Steitz, J.A., D.L. Black, V. Gerke, K.A. Parker, A. Kramer, D. Frendewey, and W. Keller. 1988. Functions of the abundant U-snRNPs. In Structure and function of major and minor SNURPS (ed. M.L. Birnstiel), pp. 115-154. Springer-Verlag, Heildelberg.

Streuli, M. and H. Saito. 1989. Regulation of tissue specific alternative splicing: Exon-specific cis-elements govern the splicing of leukocyte common antigen pre-mRNA. EMBO J. 8: 787-796.

Swanson, M.S. and G. Dreyfuss. 1988. RNA binding specificity of hnRNP proteins: A subset bind to the 3' end of the intron. EMBO J. 7: 3519-3529.

Tazi, J., C. Alibert, J. Temsamani, I. Reveillaud, G. Cathala, C. Brunel, and P. Jeanteur. 1986. A protein that specifically recognizes the 3' splice site of mammalian pre-mRNA introns is associated with a small nuclear ribonucleoprotein. Cell 47: 755- 766.

Ulfendahl, P.J., J.-P. Kreivi, and G. Skusjarvi. 1989. Role of the branch site/3' splice site region in adenovirus-1 E1A pre- mRNA alternative splicing: evidence for 5'-and 3'-splice site cooperation• Nucleic Acids Res. 3: 925-938.

Wu, J. and J.L. Manley. 1989. Mammalian pre-mRNA branch site selection by U2 snRNP involves base pairing. Genes Dev. 3: 1553-1561.

Zhuang, Y., and A.M. Weiner. 1989. A compensatory base change in U2 snRNA suppresses a branch site mutation. Genes Dev. 3: 1545-1552.

Zhuang, Y., A.M. Goldstein, and A.M. Weiner. 1989. UA- CUAAC is the preferred branch site for mammalian splicing• Proc. Natl. Acad. Sci. 86: 2752-2756.

Zeitlin, S. and A. Effstradiatis. 1984. In vivo splicing products of the rabbit ~-globin pre-mRNA. Cell 39: 589-602.

Zillman, M., S.D. Rose, and S.M. Berget. 1987. U1 small nuclear ribonucleoproteins are required early during spliceosome assembly. Mol. Cell. Biol. 7: 2877-2883.





10.1101/gad.3.12b.2113Access the most recent version at doi: 3:1989, Genes Dev.

R Reed The organization of 3' splice-site sequences in mammalian introns.

References

http://genesdev.cshlp.org/content/3/12b/2113.full.html#ref-list-1

This article cites 51 articles, 20 of which can be accessed free at:

License

ServiceEmail Alerting

click here.right corner of the article or

Receive free email alerts when new articles cite this article - sign up in the box at the top

Copyright © Cold Spring Harbor Laboratory Press


http://genesdev.cshlp.org/lookup/doi/10.1101/gad.3.12b.2113

http://genesdev.cshlp.org/content/3/12b/2113.full.html#ref-list-1

http://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=protocols;10.1101/gad.3.12b.2113&return_type=article&return_url=http://genesdev.cshlp.org/content/10.1101/gad.3.12b.2113.full.pdf

http://genesdev.cshlp.org/cgi/adclick/?ad=55564&adclick=true&url=https%3A%2F%2Fhorizondiscovery.com%2Fen%2Fcustom-synthesis%2Fcustom-rna%3Futm_source%3DCSHL_RNA%26utm_medium%3Dbanner%26utm_campaign%3Dcustom_synth%26utm_term%3Doligos%26utm_content%3Djan21



the organization of 3' splice-site sequences in mammalian...

Documents