Cloning of Caenorhabditis U2AF⁶⁵: an Alternatively Spliced RNA Containing a Novel Exon

Abstract

The U2 small nuclear ribonucleoprotein particle (snRNP) auxiliary factor, U2AF, is an essential splicing factor required for recognition of the polypyrimidine tract and subsequent U2 snRNP assembly at the branch point. Because Caenorhabditis elegans introns lack both polypyrimidine tract and branch point consensus sequences but have a very highly conserved UUUUCAG/R consensus at their 3 splice sites, we hypothesized that U2AF might serve to recognize this sequence and thus promote intron recognition in C. elegans. Here we report the cloning of the gene for the large subunit of U2AF, uaf-1. Three classes of cDNA were identified. In the most abundant class the open reading frame is similar to that for the U2AF 65 from mammals and flies. The remaining two classes result from an alternative splicing event in which an exon containing an in-frame stop codon is inserted near the beginning of the second RNA recognition motif. However, this alternative mRNA is apparently not translated. Interestingly, the inserted exon contains 10 matches to the 3 splice site consensus. To determine whether this feature is conserved, we sequenced uaf-1 from the related nematode Caenorhabditis briggsae. It is composed of six exons, including an alternatively spliced third exon interrupting the gene at the same location as in C. elegans. uaf-1 is contained in an operon with the rab-18 gene in both species. Although the alternative exons from the two species are not highly conserved and would not encode related polypeptides, the C. briggsae alternative exon has 18 matches to the 3 splice site consensus. We hypothesize that the array of 3 splice site-like sequences in the pre-mRNA and alternatively spliced exon may have a regulatory role. The alternatively spliced RNA accumulates at high levels following starvation, suggesting that this RNA may represent an adaption for reducing U2AF 65 levels when pre-mRNA levels are low.Introns are recognized by components of the spliceosome at specific sequences near their 5Ј and 3Ј ends that are highly conserved within a species (13). Almost all introns have good matches to these consensus sequences. In Saccharomyces cerevisiae, for example, the site at which the U2 small nuclear ribonucleoprotein particle (snRNP) base pairs with the intron to define the branch point is a perfectly conserved UACUAAC (15, 31). In contrast, mammals and flies have only a relatively weak consensus at this site. A second consensus sequence exists in all organisms at the site at which the U1 snRNP base pairs with the 5Ј splice site (19,20,38). In mammals, the 3Ј splice site appears to be defined by a polypyrimidine tract upstream of the 3Ј end and by an AG/R at the intron-exon boundary (13). The polypyrimidine tract serves as the binding site for essential splicing factor U2AF, which is thought to increase the affinity of U2 for the branch point (17). The 3Ј splice site itself is recognized by a still-undefined component of the spliceosome.In Caenorhabditis elegans, introns tend to be quite short (approximately 50 bp), althoug...

Caspase-2 pre-mRNA alternative splicing: Identification of an intronic element containing a decoy 3' acceptor site

“…In the Rous sarcoma virus system, a negative regulator of splicing element can interact with a downstream 3Ј splice site (39). In Caenorhabditis, an alternative exon in the U2AF 65 pre-mRNA was found to contain multiple copies of the worm 3Ј splice site consensus sequence (40). Similar to what we observed in this study, these sites, although seemingly normal, are not used as 3Ј splice sites.…”

Molecular genetic analysis of U2AF59 in Schizosaccharomyces pombe: Differential sensitivity of introns to mutational inactivation

“…At the outset of our work on U2AF 59 , we considered it entirely possible that splicing of the many fission yeast introns lacking an extended run of pyrimidines between the branchpoint and 39 splice site might not require this splicing factor+ Precedent for such a scenario is set by the finding that two S. cerevisiae splicing factors, Slu7p and Prp18p, are required for splicing only of introns with long branchpoint to 39 splice site distances (Jones et al+, 1995;Zhang & Schwer, 1997)+ Counter to this notion, we find that splicing of a diverse array (and thus, more than likely, all) introns in this organism is inhibited upon mutational inactivation of U2AF 59 + Moreover, the requirement for functional U2AF 59 to splice cdc2-Int2 persisted even after replacement of its pyrimidine tract with purines+ Taken together, these observations render unlikely the possibility that a distinct factor with a different binding specificity replaces U2AF 59 during splicing of S. pombe pre-mRNAs lacking strong polypyrimidine tracts+ Rather, features other than, or in addition to, the 39 polypyrimidine tract must facilitate recruitment of this protein to the substrate+ One possibility is that fission yeast U2AF 59 also recognizes the 39 AG dinucleotide that, like the polypyrimidine tract, is required prior to the first transesterification reaction in this organism (Romfo & Wise, 1997)+ Consistent with this idea, an early study with partially purified human U2AF indicated a preference for substrates containing an intact 39 splice site (Ruskin et al+, 1988)+ In addition, there is evidence to suggest that the short polypyrimidine tract and 39 splice site are recognized as a unit by U2AF 65 in C. elegans (Zorio et al+, 1997)+ An alternative, but not exclusive, scenario is that U2AF 59 does not directly contact the 39 ends of introns lacking a strong polypyrimidine tract, but rather is tethered to the pre-spliceosome by interactions with other splicing factors such as SF1/BBP (see Introduction)+ This notion is also appealing for S. cerevisiae Mud2p, which displays conservation of the pseudo-RRM that is involved in protein-protein interactions, but not of the two classical RRMs that resemble proteins known to directly contact RNA (Abovich et al+, 1994)+ The lack of a role for sequences downstream from the branchpoint prior to the first transesterification reaction in S. cerevisiae (Patterson & Guthrie, 1991;Rymond & Rosbash, 1992) is in keeping with this model+ While all splicing events examined here appear to have at least some requirement for functional U2AF 59 , our data also suggest that a polypyrimidine tract is relevant to the participation of this protein in splicing of pre-mRNAs in vivo+ In particular, the second intron of the cdc2 gene, which contains exclusively pyrimidines between the branchpoint and 39 splice site, displayed the most dramatic splicing defect of any intron tested in both temperature-sensitive mutants analyzed (Figs+ 4, 5, and 6)+ We have shown elsewhere that the relatively inefficient splicing of wild-type cdc2-Int2 is a consequence of several factors including its relatively large size for this organism (C+M+ Romfo, C+J+ Alvarez, & J+A+ Wise, in prep+), the presence of a stem-loop structure encompassing the 59 splice site (Alvarez et al+, 1996)…”

“…U2 auxiliary factor (U2AF), a heterodimer composed of 65-and 35-kDa subunits (Zamore & Green, 1989), was originally identified by biochemical complementation as a component of mammalian splicing extracts that promotes stable association of the U2 snRNP with the branchpoint (Ruskin et al+, 1988)+ The large subunit is composed of five distinct domains: an amino-terminal segment rich in Arg-Ser dipeptides, followed by a short hinge region responsible for dimerization with the small subunit, and a large carboxy terminus consisting of two standard RNA recognition motifs (RRMs) and a third RRM-like module sometimes referred to as a pseudo-RRM (Zamore et al+, 1992;Birney et al+, 1993)+ Orthologs of U2AF 65 have now been identified in a variety of organisms, including budding yeast (Abovich et al+, 1994), fission yeast (Potashkin et al+, 1993), Drosophila (Kanaar et al+, 1993), and Caenorhabditis elegans (Zorio et al+, 1997)+ U2AF large-subunit function has been extensively analyzed in vitro (e+g+, Ruskin et al+, 1988;Zamore & Green, 1991;Zamore et al+, 1992;Valcarcel et al+, 1993Valcarcel et al+, , 1996Gaur et al+, 1995;Singh et al+, 1995;Zuo & Maniatis, 1996;reviewed in Krämer, 1996), while only limited information is as yet available regarding its function in vivo (Kanaar et al+, 1993;Potashkin et al+, 1993;Rudner et al+, 1998aRudner et al+, , 1998b)+ Splicing activity can be restored to depleted human extracts by the addition of either human U2AF 65 or Drosophila U2AF 50 , indicating that the function as well as the structure of this factor is conserved among metazoa (Ruskin et al+, 1988;Zamore & Green, 1991)+ Consistent with the in vitro data, chromosomal deletions encompassing the gene encoding Drosophila U2AF 50 result in embryonic lethality (Kanaar et al+, 1993), and RNA interference experiments indicate that blocking production of the C. elegans large subunit is also lethal (T+ Blumenthal, pers+ comm+)+ However, it has not been possible to show directly that loss of large subunit function in either flies or worms leads to a splicing defect (D+ Rudner and D+ Rio, unpubl+ observations cited in Rudner et al+, 1998aRudner et al+, , 1998b; T+ Blumenthal, pers+ comm+)+ A gene en...…”

Conservation of functional domains involved in RNA binding and protein–protein interactions in human and Saccharomyces cerevisiae pre-mRNA splicing factor SF1

“…35 (Wu & Maniatis, 1993;Kohtz et al+, 1994;Staknis & Reed, 1994)+ In yeast, U1 snRNP binds to the 59 splice site during the formation of CC1 (Séraphin & Rosbash, 1989)+ CC2 contains, in addition to U1 snRNP, Mud2p, a protein identified in a screen for synthetic lethality with a mutation in U1 snRNA (Abovich et al+, 1994)+ Mud2p is not tightly associated with U1 snRNP and contacts the pre-mRNA directly in the presence of a functional branch site sequence+ In addition, Mud2p is directly or indirectly involved in the recognition of the U residue immediately upstream of the branch site (Rain & Legrain, 1997)+ Amino acid sequence comparisons suggested that Mud2p is distantly related to U2AF 65 , which contains an N-terminal arginine-serine-rich (RS) domain and three RNA recognition motifs (RRMs, see Fig+ 1; Zamore et al+, 1992;Abovich et al+, 1994)+ The first two RRMs are consensus RRMs, whereas the third one is degenerate (Birney et al+, 1993)+ These features are conserved in U2AF 65 homologues in Drosophila, Caenorhabditis elegans, and Schizosaccharomyces pombe (Kanaar et al+, 1993;Potashkin et al+, 1993;Zorio et al+, 1997)+ Mud2p lacks an RS domain and shows highest homology to U2AF 65 in the region of the degenerate RRM3 (31% identity and 49% similarity, see Fig+ 1; Abovich et al+, 1994)+ We have previously purified human splicing factor SF1 (hSF1) based on its activity in the formation of pre-splicing complex A (Krämer, 1992)+ This complex forms by binding of U2 snRNP and associated proteins to complex E and involves base pairing of U2 snRNA with the branch site (reviewed in Hodges & Beggs, 1994;Reed, 1996)+ Inspection of the amino acid sequence of hSF1 (Arning et al+, 1996) revealed the presence of a maxi-KH domain, a motif found in a variety of proteins associated with RNA (Fig+ 1; Siomi et al+, 1993a;Musco et al+, 1996), and a zinc knuckle (of the consensus sequence Cys-X 2 -Cys-X 4 -His-X 4 -Cys), which is a characteristic motif in retroviral nucleocapsid proteins (Darlix et al+, 1995)+ Consistent with the presence of these motifs, hSF1 was shown to bind RNA in a sequence-independent fashion, although a preference for binding to guanosine or uridine-rich RNA was observed (Arning et al+, 1996)+ The C-terminal half of SF1 is rich in proline residues and recently has been shown to be the target for binding of form...…”