Activation of a cryptic TACTAAC box in the Saccharomyces cerevisiae actin intron.

Abstract

We constructed a translational fusion between the Saccharomyces cerevisiae actin gene and the Escherichia coli (-galactosidase structural gene such that expression of 0-galactosidase activity required accurate splicing of the actin intron. Using this chimeric gene, we generated a series of internal deletions which removed the TACTAAC box or, in addition, TACTAAC-like sequences within the intron. Analysis of the fusion transcripts produced in these deletions allowed us to conclude that the TACTAAC-like sequence TACTAAG can substitute, albeit inefficiently, for the authentic TACTAAC box in the splicing process. These results indicate that the yeast splicing machinery can utilize a cryptic TACTAAC box, but there are requirements for primary sequence and proper position.A key step in the splicing process is the definition of the precise boundaries of the intron by the splicing machinery. In the yeast Saccharomyces cerevisiae, a subset of nuclear genes encoding mRNAs contain intervening sequences (IVS) (34). Previous work has identified two regions within the intron which are required for splicing. The sequence at the 5' intron-exon junction, although subject to substantial variation about a consensus sequence in metazoans, in yeasts is an essentially 100%o conserved hexanucleotide (11). This apparent lack of variance in yeast splicing signals allowed the identification of an additional conserved sequence within the intron, the so-called TACTAAC box, a heptanucleotide found in all yeast introns 4 to 53 nucleotides upstream of the AG at the 3' splice junction. The importance of these conserved sequences in the splicing process was demonstrated by the observation that deletion of either the 5' consensus sequence (4) or the TACTAAC box (11,12,26) abolishes splicing and leads to the accumulation of fulllength precursor (11,12,26).While consensus sequences at the 5' (GTAAGT) and 3' [(PY)nAG] intron-exon junctions had been recognized in metazoan introns (1, 21), there has been no evidence to suggest an analog of the TACTAAC box. This apparent difference has recently been verified with the development of extracts which carry out the splicing reaction in vitro. A detailed analysis of the transcripts in both mammalian and yeast extracts (22,24), and more recently the characterization of intermediates in the splicing process in vivo (3,28,37), has shown that splicing of pre-mRNA in both systems occurs via a common intermediate. This intermediate has been termed a lariat, because the 5' end of the intron is joined by a 2'-5' phosphodiester bond to a site within the intron. The location of this 2'-5' linkage, or branch (36), is within the TACTAAC box in yeasts (3). In mammalian cells, branch points are determined primarily by their location with respect to the 3' splice site regardless of the sequence surrounding the branch point, and an obvious consensus sequence is less readily identifiable (9,24,27,30,31).The novel nature of these intermediates argues that splicing must proceed by fundamentally similar mechanisms in * Corresp...

Effects on mRNA splicing of mutations in the 3' region of the Saccharomyces cerevisiae actin intron.

“…1). Previously we demonstrated that the same sequence at the normal ICS position still allows fairly efficient splicing (9), whereas others have shown that this upstream cryptic TACTAAG is used for branch formation, albeit inefficiently, only if the actin ICS is deleted (6) or contains a C-to-A transversion at the third position (39). Other ICS mutations prevent use of the upstream TACTAAG (9,39).…”

“…Other ICS mutations prevent use of the upstream TACTAAG (9,39). This inefficient use of the upstream TACTAAG sequence (6,39), as compared with the relatively efficient use of a TACTAAG mutation at the ICS (9), suggests that subtle changes in sequence context may effect the recognition of a potential branchpoint. The data of Pikielny and Rosbash (28) demonstrate that alterations adjacent to either the 5' conserved junction sequence or the ICS can dramatically effect the efficiency of stage 1.…”

In vitro reconstitution of functional yeast U2 snRNPs.

“…Splicing activities of the reconstituted extracts were then assayed using wildtype (WT), A256, and A257 actin pre-mRNA substrates. All of the pre-mRNA substrates used in this experiment contained the A6 mutation, which removes a cryptic branchpoint sequence (UACUAAG) that can be used on mutation of the normal branch site:,sequence (Vijayraghavan et al 1986;Cellini et al 1986). With the WT actin pre-mRNA substrate, substantial splicing activity is seen only with the T7U2(WT) RNA.…”

Yeast pre-mRNA splicing requires a minimum distance between the 5' splice site and the internal branch acceptor site.

“…In the yeast Saccharomyces cerevyisiae, highly conserved sequences at the 5' splice site (GTATGT), 3' splice site (PyAG), and internal branch acceptor site (TACTAAC) have been documented as structural requirements for correct and efficient removal of introns from mRNA precursors (9,12,18,19,30,31,35). These highly conserved elements hatve been subject to intense study (7,11,19,24,28,38), whereas only little is known about the nonconserved sequences in yeast introns. Cellini et al reported that a 15-base-pair (bp) deletion between the internal branch acceptor site and the 3' splice site of the yeast actin intron has no effect on splicing (6); a deletion of 170 nucleotides which leaves intact the first 13 nucleotides at the 5' splice junction and most of the 3' part of this intron also does not adversely affect splicing (12).…”