| Nucleic Acids Research | Pages |
Different positioning elements select poly(A) sites at the3[prime]-end of GCN4 mRNA in the yeast Saccharomyces cerevisiae
Introduction
Materials And Methods
Yeast strains
Plasmid construction and cloning
RNA analysis
RACE (rapid amplification of cDNA ends)
Results
Point mutations in an AAGAA motif result in a loss of specificity of poly(A) site selection
A temperature-sensitive form of the processing factor Rna15p causes an alteration in poly(A) site selection in the wild-type GCN4 3[prime]-UTR similar to the AAGAA mutation GCN4mut-p
The GCN4 3[prime]-end comprises additional positioning motifs for poly(A) sites which resemble putative efficiency elements
Expression of the GCN4mut-3 allele in an rna15-2 strain results in a novel pattern of selected poly(A) sites
Discussion
Acknowledgements
References
Different positioning elements select poly(A) sites at the3[prime]-end of GCN4 mRNA in the yeast Saccharomyces cerevisiae
Received August 19, 1999; Revised and Accepted October 28, 1999
ABSTRACT Cleavage and polyadenylation of eukaryotic mRNA requires efficiency and positioning elements in the 3[prime]-untranslated region (3[prime]-UTR) of the mRNA. Specific point mutations were introduced into the yeast GCN4 3[prime]-UTR to detect sequence motifs which are involved in the positioning of the poly(A) site. 3[prime]-End proces-sing activities of different GCN4 3[prime]-UTR alleles were measured in an in vivo test system. Point mutations in an AAGAA motif defocussed selection of the poly(A) sites of the GCN4 3[prime]-UTR to various additional poly(A) sites instead of the single site of the wild-type GCN4 3[prime]-UTR. A strain with an intact wild-type GCN4 3[prime]-UTR but impaired in RNA15 encoding an RNA-binding processing factor showed a similar defocussed pattern of poly(A) site selection. Remarkably, two additional sequence motifs upstream of the AAGAA motif which resemble yeast efficiency motifs independently affected poly(A) site positioning but not efficiency of 3[prime]-end processing. Mutations in one motif resulted in an additional upstream poly(A) site. Alterations of the other motif shifted the poly(A) sites exclusively to two downstream poly(A) sites. These data suggest several contact points between the precursor mRNA and the polyadenylation machinery in yeast.
INTRODUCTION
Most eukaryotic mRNAs carry a poly(A) tail which is added to specific cleavage sites in the 3[prime]-untranslated region (3[prime]-UTR) in a post-transcriptional process (for recent reviews see 1,2). The sequence requirements for efficient 3[prime]-end processing are well defined in higher eukaryotes. A highly conserved AAUAAA motif is located upstream and a GU-rich region downstream of the poly(A) site. The positioning of the poly(A) site is determined by the distance between these two sequence motifs. In yeast, less conserved sequence motifs vary in their arrangement. Recently, a computer alignment of 3[prime]-UTRs of 861 different yeast genes was performed to enlighten the question of sequence motifs (3). A model consisting of five different signal sequences was proposed based on the frequency of sequence motifs at specific positions in the 3[prime]-UTRs. Three sequence motifs are located upstream of the poly(A) site which itself is one of the five elements and most often found to be a PyA dinucleotide. The far upstream efficiency motif and the positioning motif located further downstream were previously confirmed by experimental data (4,5). The computer analysis revealed an additional T-rich region in close proximity upstream of the poly(A) site and a motif localized downstream of the poly(A) site. However, experimental studies of different yeast 3[prime]-UTRs showed that each region has its unique features, including specific sequence motifs and their peculiar arrangement in the 3[prime]-UTR (4,6-9). The yeast CYC1 gene contains one of the best-studied 3[prime]-ends with independently functioning efficiency and positioning motifs. The AAGAA positioning motif of the CYC1 3[prime]-UTR can be deleted without affecting the processing efficiency (10). A total mutagenesis revealed AATAAA and AAAAAA as optimal positioning motifs (11). An (A)11 motif directs the positioning of poly(A) sites of the FBP1 gene (9). In the case of the FBP1 3[prime]-UTR the positioning motif is located upstream of the efficiency motif. Deletion of the positioning motif results in a more scattered pattern of poly(A) sites due to additional downstream sites (9). An entirely different positioning motif was found in the TRP4 3[prime]-UTR. Its sequence composition, TAGT, resembles efficiency motifs of other yeast genes. Correspondingly, a point mutation within this motif not only altered poly(A) site selection but also drastically reduced the 3[prime]-end processing activity (12). An additional difference to the formerly described positioning motifs was the localization of new poly(A) sites exclusively upstream of the wild-type poly(A) site due to a point mutation in the TAGT motif.
Besides the cis sequences, a complex protein machinery is necessary for mRNA 3[prime]-end processing in eukaryotes. There are at least three different complexes in yeast, each consisting of several protein factors which are involved in this process. They are called PFI for polyadenylation factor I, and CFI and CFII (cleavage factors I and II). An additional factor, which might also be part of the PFI complex, the poly(A) polymerase Pap1p, has been described (reviewed in 1,13). Rna15p is part of the CFI complex and is a quite well-studied processing factor. It binds RNA with a preference for a U-rich consensus sequence deduced from in vitro selection experiments (SELEX) (14,15). In strains carrying a temperature-sensitive allele of RNA15, an alteration in poly(A) site selection could be observed at the restrictive temperature. The ACT1 3[prime]-end, which regularly uses five different poly(A) sites, was taken as example. The temperature shift led to alterations in the frequency with which each poly(A) site was used (16).
In this study, the positioning motifs of the complex GCN4 3[prime]-UTR were analysed. A former deletion analysis of the GCN4 3[prime]-UTR suggested two TTTTTAT sequences as efficiency motifs for 3[prime]-end processing (7). The region in between these two sequence motifs was also necessary for full processing activity. There are several TA-rich sequence motifs located in this region which resemble efficiency motifs of other genes. In addition, a deletion of a part of this region altered selection of the poly(A) site (7). Since an additional AAGAA motif, a putative positioning element which has not been analysed so far, is also present in this region, we wondered whether the 3[prime]-UTR of GCN4 contains several positioning elements. Here, we show that alterations in the AAGAA sequence in cis resulted in similar defocussed mRNA 3[prime]-ends as found with an intact GCN4 3[prime]-UTR and defects in Rna15p function in trans. In addition, positioning of the GCN4 3[prime]-end could also be changed in the presence of an intact AAGAA by introducing single point mutations elsewhere in the GCN4 3[prime]-UTR. We found point mutations which resulted in shorter as well as longer mRNA species. Therefore, we propose the existence of several contact points between the precursor mRNA and the polyadenylation machinery.
MATERIALS AND METHODS
Yeast strains
The in vivo test system vectors carrying different alleles of the GCN4 3[prime]-UTR were transformed into the yeast strain RH1631 (MATa, ura3-52) which is a derivative of the Saccharomyces cerevisiae laboratory standard strain X2180-1A (MATa gal2 SUC2 mal CUP1). The test system cassette was integrated at the URA3 locus. The following strains were derived from RH1631: RH2258 (ura3-52::pME1566) carrying GCN4mut-1; RH2259 (ura3-52::pME1567) with GCN4mut-2; RH2260 (ura3-52::pME1568) with GCN4mut-3; RH2261 (ura3-52::pME1569) with GCN4mut-4; RH2262 (ura3-52::pME1570) with GCN4mut-p. The strains RH2395 (MATa, ura3-1::pME800, trp1-1, ade2-1, leu2-3,112, his3-11,15, rna15-2), carrying the wild-type GCN4, and RH2411 (MATa, ura3-1::pME1568, trp1-1, ade2-1, leu2-3,112, his3-11,15, rna15-2) were based on rna15-2W (MATa, ura3-1, trp1-1, ade2-1, leu2-3,112, his3-11,15, rna15-2), kindly provided by Dr F. Lacroute (Centre National de la Recherche Scientifique, Gif-sur-Yvette, France). Correct integration at the URA3 locus was examined by Southern analysis (17). The yeast strains RH1712 (MATa, ura3-52::pME800), carrying the wild-type GCN4 3[prime]-UTR in positive orientation, and RH1713 (MATa, ura3-52::pME801), with the GCN4 3[prime]-UTR inserted in negative orientation in the in vivo test system (7), were used as controls. Yeast strains were cultivated on YEPD medium or on SD minimal medium with appropriate supplements at 30°C (18). Yeast rna15-2W derivatives were grown at 25°C as the permissive temperature and shifted to 37°C as the restrictive temperature to induce the temperature-sensitive defect of the rna15-2 allele.
Plasmid construction and cloning
The different point mutations were introduced into the GCN4 3[prime]-UTR by site-directed PCR mutagenesis using the oligonucleotides GCN4K-1B (5[prime]-CATTCACTTAGATACGTATTTAATCTTTTTTATG-3[prime]) to generate GCN4mut-p, GCN4K-7 (5[prime]-CTATTTCGTTATGCACGAGAATG-3[prime]) to construct GCN4mut-1, GCN4K-2 (5[prime]-CTAAGTGAATGTATCCGTA-TCGTTATACACG-3[prime]) for GCN4mut-2, GCN4K-3 (5[prime]-CTT-ATCTAAGTGAACGTCTCTATTTCGTTATAC-3[prime]) for GCN4-mut-3, and GCN4K-9 (5[prime]-CATTCACTCAGATAAGAATTTAATC-3[prime]) for GCN4mut-4. The oligonucleotides OLCE3 (5[prime]-CGCGGGTACCCTTGGAAAATGAGGTTGCCAG-3[prime]), carrying a KpnI site, and OLCE6 (5[prime]-GCGCAGATCTCCCC-CATCGTGAGTG-3[prime]), with a BglII site, were used as flanking primers. The fragments were cloned into the KpnI and BglII sites of the multiple cloning site of the test system of plasmid pME798. The resulting plasmids were pME1570 (GCN4mut-p), pME1566 (GCN4mut-1), pME1567 (GCN4mut-2), pME1568 (GCN4mut-3) and pME1569 (GCN4mut-4). The plasmid pME800 carries the wild-type GCN4 3[prime]-UTR in the in vivo test system (7).
RNA analysis
Total RNA from S.cerevisiae was isolated as described (19). For northern analysis 20 µg of total RNA were separated on a formaldehyde-agarose gel, transferred to a positively charged nylon membrane (Biodyne B; PALL) by electroblotting and hybridized with 32P-labelled DNA fragments as described previously (7). The DNA fragments were random labelled using the HexaLabel DNA Labelling Kit from MBI Fermentas. Band intensities were visualized by autoradiography and quantified using a BAS-1500 Phosphorimaging scanner (Fuji).
RACE (rapid amplification of cDNA ends)
RACE experiments were performed based on the description in Frohman et al. (20). Aliquots of 100 ng of total RNA of the different S.cerevisiae strains were reverse transcribed using the M-MuLV reverse transcriptase (MBI Fermentas) with a poly(dT)18 primer. In a following PCR reaction 3[prime]-ends of the different GCN4 test system transcripts were amplified with the test system-specific primer OLCE9 (5[prime]-GGTCAATCTTTGTTAAAGAATAGG-3[prime]). OLCE9 binds to an intron sequence which is spliced out in the ACT1 wild-type mRNA, but which is still present in the transcripts of the in vivo test system, where the 3[prime]-splice site is missing. DNA fragments were isolated and cloned into pBluescript SK+ (Stratagene). The poly(A) site was determined by sequencing using the T7 Sequencing Kit (Pharmacia).
RESULTS
The GCN4 3[prime]-UTR has previously been operationally divided into five different regions, EL1-EL5 (Fig. 1). The different point mutations in this study were located in regions EL2, EL3 and EL4 between the two TTTTTAT motifs which were shown to be necessary for efficient 3[prime]-end processing (7). The different alleles of the GCN4 3[prime]-UTR were cloned into an in vivo test system consisting of the ACT1 promoter and the ADH1 3[prime]-UTR (Fig. 2A). The ADH1 3[prime]-UTR is able to process all transcripts which are initiated at the strong ACT1 promoter. In between, the different alleles of the GCN4 3[prime]-UTR were cloned (Fig. 2A). By using the in vivo test system the 3[prime]-end processing efficiency of the investigated 3[prime]-UTR can be determined. For this purpose the amount of transcripts processed in the GCN4 3[prime]-UTR is compared to the total amount of transcripts initiated at the ACT1 promoter. The test systems containing the different GCN4 3[prime]-UTR alleles were integrated into the yeast genome.
Figure 1. Sequence of the GCN4 3[prime]-UTR. The region had previously been divided into five different parts (EL1-EL5) as a result of a deletion analysis (7). The different point mutations which were introduced into the GCN4 3[prime]-UTR are shown below. The exchanged nucleotides are underlined in the sequence and the resulting nucleotides of the sense strand are noted. The alleles carrying different point mutations were named GCN4mut-p for alteration of the putative positioning motif AAGAA and GCN4mut-1-GCN4mut-4 for mutations in other sequence motifs according to their 5[prime]->3[prime] succession in the 3[prime]-UTR. The putative sequence motifs in which the point mutations were located are underlined in dark grey for AAGAA and light grey rectangles, respectively. The efficiency motifs of the GCN4 3[prime]-UTR were described ealier (7) and are marked by black rectangles. Dark points above the sequence indicate intervals of 10 nt.
Figure 2. (A) Schematic presentation of the in vivo test system which was used to investigate the 3[prime]-end processing activity of the GCN4 3[prime]-UTR. It consists of the promoter and a small part of the open reading frame of the ACT1 gene and the ADH1 3[prime]-UTR with a multiple cloning site in between (diagonal stripes) (7). The different GCN4 3[prime]-UTRs were cloned into this multiple cloning site. The resulting transcripts are indicated below. The sizes of the read-through and the truncated transcripts are ~720 and 380 nt, respectively. (B) Processing efficiency of the GCN4 alleles carrying point mutations in the 3[prime]-UTR. The processing efficiency was determined by northern analysis of 20 µg total RNA, using a 460 bp MluI-KpnI fragment of the 5[prime]-region of the ACT1 gene as probe. The upper band of the resulting three bands represents the transcript of the genomic ACT1 gene. The band in the middle shows a read-through transcript which is processed in the ADH1 3[prime]-UTR of the test construct. The lower band represents truncated transcripts of the test constructs which were processed in the GCN4 3[prime]-UTR. The wild-type GCN4 3[prime]-UTR in positive, functional or negative, non-functional orientation served as controls. Quantification was performed by phosphorimager analysis. The processing activities are expressed by the percentage of truncated transcripts compared to the sum of truncated and read-through transcripts. The columns represent northern hybridizations of at least three independent RNA isolations. The standard deviations did not exceed 15%.
Point mutations in an AAGAA motif result in a loss of specificity of poly(A) site selection
The 3[prime]-ends of yeast mRNAs include positioning as well as efficiency elements of mRNA 3[prime]-end processing. Our aim was to analyse sequence motifs in the GCN4 3[prime]-UTR for their possible involvement in poly(A) site positioning. AAGAA and similar motifs have been shown to influence selection of the poly(A) site of other yeast genes, e.g. CYC1 (10). The GCN4 3[prime]-UTR contains an AAGAA motif which is located at position +73 relative to the translational stop codon. To investigate its possible role in poly(A) site selection, specific point mutations were introduced resulting in an ACGTA sequence instead of AAGAA (Fig. 1). The allele was named GCN4mut-p (p stands for positioning). The processing efficiency of the GCN4mut-p allele was determined by northern analysis experiments using the 5[prime]-part of the ACT1 gene as probe. This probe hybridizes to all transcripts of the in vivo test system, as well as to the ACT1 mRNA. The point mutations in AAGAA revealed similar amounts of truncated transcripts processed in the GCN4 3[prime]-UTR as the wild-type allele. Therefore, these point mutations had no significant effect on the processing efficiency (Fig. 2B). The wild-type GCN4 3[prime]-UTR in negative orientation as an additional control was not functional and revealed a processing efficiency of <5%.
The poly(A) sites of GCN4mut-p were localised by RACE experiments and compared to transcripts containing the wild-type GCN4 in the in vivo test system. The data represent the 3[prime]-ends of at least 20 different cDNA clones of each construct. The wild-type GCN4 in positive orientation revealed a single poly(A) site at position +96 relative to the translational stop codon. In contrast, the poly(A) sites of GCN4mut-p were found to form a scattered pattern (Fig. 3). Apart from the wild-type poly(A) site which was predominantly used (43%), additional poly(A) sites were found further downstream at positions +105, +111, +126 and +138. In the case of the poly(A) site at position +138 the preference for PyA as poly(A) site was neglected. The poly(A) tail was added to a C which was formerly followed by another C. The AAGAA motif is therefore a GCN4 3[prime]-UTR positioning element which might be responsible for the use of a single specific poly(A) site.
Figure 3. Schematic presentation of the poly(A) site selection of the wild-type GCN4 3[prime]-UTR and the alleles carrying different point mutations. The poly(A) sites were determined by RACE experiments. The TTTTTAT sequence corresponds to the efficiency motif located more downstream. The poly(A) sites are indicated by arrows, as are the dinucleotides where cleavage occurred. The positions are relative to the translational stop codon. The frequency with which the poly(A) site was chosen is shown in parentheses and indicated by the thickness of the arrow. Each arrow represents at least two independent cDNA clones. Altogether 20 cDNA clones of each test construct were sequenced, with the exception of the GCN4mut-p allele, for which 30 clones were used.
A temperature-sensitive form of the processing factor Rna15p causes an alteration in poly(A) site selection in the wild-type GCN4 3[prime]-UTR similar to the AAGAA mutation GCN4mut-p
The 3[prime]-ends of yeast precursor mRNAs are recognised by the processing machinery which is a multisubunit protein complex. The protein factor Rna15p represents a subunit of the processing machinery which directly binds to RNA (14) and which is suggested to be involved in poly(A) site selection (16). Therefore, we wondered how defects in the corresponding RNA15 gene would affect poly(A) site selection of the GCN4-containing test system transcript.
The in vivo test system carrying the wild-type GCN4 3[prime]-UTR was transformed into a yeast strain carrying an rna15-2 allele resulting in a temperature-sensitive gene product (14). Total RNA was isolated from cultures grown at the permissive temperature (25°C) and 15, 30 and 60 min after shifting to the non-permissive temperature (37°C). The northern analysis which demonstrates the processing efficiency can be seen in Figure 4. The truncated test system transcript decreased by a factor of three after 60 min incubation at 37°C. Additionally, more read-through transcripts were synthesized and therefore a change in the ratio between read-through and truncated transcripts can be observed. After 60 min incubation at the non-permissive temperature the processing activity of the GCN4 3[prime]-UTR is decreased to 81 compared to 97% at the permissive temperature. Rna15p seems to have an auxiliary function in efficient processing of the precursor mRNA.
Figure 4. Processing activity of the test construct carrying a wild-type GCN4 3[prime]-UTR in a strain with the rna15-2 temperature-sensitive mutant allele. The cultures were grown at 25°C and then shifted to 37°C. Total RNA was isolated prior to the temperature shift (0) and 15, 30 and 60 min after it. Aliquots of 20 µg of RNA from each sample equalized by the extinction at a wavelength of 260 nm were hybridized with a probe from the 5[prime]-part of the ACT1 gene as described in Figure 2B. The processing activities show the amount of truncated transcripts compared to the total amount of all transcripts expressed by the in vivo test system. Quantification of the decrease in total amounts of truncated transcripts is shown below. The amount of truncated transcripts prior to the temperature shift (0 min) was set as 100%. The columns represent an average of three independent measurements and the standard deviation did not exceed 15%.
To determine alterations in poly(A) site selection, RACE experiments were performed for the non-shifted culture and the culture which was grown for 60 min at 37°C, respectively. The test system RNA of the non-shifted culture exhibits a single poly(A) site at position +96 relative to the translational stop codon, which is identical to the wild-type yeast strain carrying an intact RNA15 allele (Fig. 3). However, the transcripts of the temperature-shifted cells revealed additional poly(A) sites further downstream (Fig. 5A). The major poly(A) site was located at the wild-type position (+96) and was chosen in 60% of all investigated test system cDNAs. Additional poly(A) sites were located at positions +102, +105, +111, +132 and +137, respectively. Each of these poly(A) sites was used at a much lower frequency than the poly(A) site at position +96. The preference for PyA is not strictly followed since CT and GC were also chosen as poly(A) sites. The positions of the new poly(A) sites of the intact GCN4 3[prime]-UTR in an rna15 mutant strain were similar to the pattern and the reduced sequence specificity which were obtained by the point mutated AAGAA motif in a wild-type RNA15 genetic background. These data suggest that Rna15p interacts with the AAGAA motif at position +73 of the GCN4 3[prime]-UTR.
Figure 5. (A) Schematic presentation of poly(A) site selection of transcripts resulting from the in vivo test system carrying the wild-type GCN4 3[prime]-UTR in an rna15-2 strain. The poly(A) sites were detected by RACE experiments of RNA from cultures grown at the permissive temperature (25°C) and cultures grown for 60 min at the restrictive temperature (37°C). Poly(A) sites are indicated by arrows, as are the dinucleotide sequences where cleavage occurs. The position of the poly(A) site relative to the translational stop codon is shown below and the relative frequency of these mRNA ends in parentheses above the arrows. The data represent 30 independent clones of three indepent RNA isolations for each time point. (B) Schematic presentation of poly(A) sites which were used in the GCN4mut-3 allele in an rna15-2 background. Determination and presentation of poly(A) sites correspond to (A).
The GCN4 3[prime]-end comprises additional positioning motifs for poly(A) sites which resemble putative efficiency elements
Changes in the AAGAA motif or function of the Rna15 protein defocus 40-60% of the 3[prime]-ends of the GCN4 mRNA. The remaining 3[prime]-ends are still polyadenylated at the same position as in the presence of the wild-type GCN4 sequence and intact Rna15p. Therefore, we expected that additional positioning motifs might be present in the GCN4 3[prime]-UTR. A single point mutation in a TAGT motif changed efficiency of 3[prime]-end processing as well as poly(A) site selection of the TRP4 3[prime]-UTR (12). In between the two TTTTTAT motifs of the GCN4 3[prime]-UTR, which are necessary for efficient 3[prime]-end processing (7), there are different variations of TA(T)GT and TATATA motifs, which are the common efficiency motifs in many yeast genes. Different point mutations were introduced into these sequence motifs to determine if there is any influence of these degenerate motifs on poly(A) site selection. Selected sequence motifs in the GCN4 3[prime]-UTR were TGTATA, starting at position +42 relative to the translational stop codon, TAGATACA at position +54 and TAGAT at position +68. The point mutations were numbered according to their 5[prime]->3[prime] appearance in the GCN4 3[prime]-UTR resulting in GCN4mut-1 for the point mutation in TGTATA. GCN4mut-2 and GCN4mut-3 were located in the TAGATACA motif and GCN4mut-4 in the TAGAT motif.
Our first objective was to test by northern analysis whether one of these point mutations, which were inserted in putative efficiency motifs, changes the rate of RNA processing (Fig. 2B). However, the efficiency of 3[prime]-end processing was not significantly affected by any of the point mutations in the sequence motifs. They all showed >85% transcripts which were processed in the GCN4 3[prime]-UTR. Therefore, these sequences are not significantly involved in efficient RNA precursor processing.
To determine the poly(A) sites of the different GCN4 test system transcripts RACE experiments were performed (Fig. 3). GCN4mut-1 and GCN4mut-2 revealed only the wild-type poly(A) site at position +96. These two alterations of the GCN4 3[prime]-UTR therefore had an effect neither on the selection of the poly(A) site nor on the efficiency of 3[prime]-end processing.
The point mutation GCN4mut-3 changing the TAGATACA motif to TAGAGACG revealed two poly(A) sites located further downstream (Fig. 3). The wild-type poly(A) site was not used in this mutated allele. The major poly(A) site of this construct which occurred in 80% of all sequenced RACE products is located at a TA motif at position +105. A less frequently chosen poly(A) site was found at another TA motif at position +109. Therefore, this element seems to be important for poly(A) site selection at the wild-type site. Alternatively, a new positioning motif could have been created by this point mutation which directs cleavage at the downstream poly(A) site in a very efficient way.
In contrast, the point mutation GCN4mut-4 concerning the TAGAT motif adjacent to the putative positioning motif AAGAA resulted in the wild-type poly(A) site and an additional poly(A) site which was located further upstream at a TA motif at position +73 (Fig. 3). They were chosen with comparable frequency, the wild-type poly(A) site in 60% and the upstream poly(A) site in 40% of all cases.Processing at the novel poly(A) site separates the mutated TAGAT sequence from the AAGAA positioning motif which is located immediately downstream and adjacent to it. Therefore, changes in this element result in 3[prime]-ends which are located at the positioning element.
Expression of the GCN4mut-3 allele in an rna15-2 strain results in a novel pattern of selected poly(A) sites
The GCN4mut-3 allele was transformed into the rna15-2-carrying yeast strain to investigate the effect of non-functional Rna15p on poly(A) site selection. Test system transcripts isolated at the permissive temperature and up to 60 min after the shift to the restrictive temperature revealed similar amounts of transcripts as the wild-type GCN4 allele (data not shown). RACE experiments were performed on RNA isolated prior to and 60 min after incubation at 37°C (Fig. 5B). The downstream poly(A) site at position +105, which was also predominantly used in a wild-type RNA15 background, was found in 87% of all sequenced cDNAs in the culture grown at the permissive temperature. In contrast to the wild-type strain, the rna15 mutant strain occasionally showed the wild-type poly(A) site at position +96 even at the permissive temperature (13% of all cases). Temperature shift to 37°C resulted in an altered percentage of the use of wild-type compared to the downstream poly(A) site. After 60 min at 37°C the wild-type poly(A) site was preferred in 57% of all sequenced cDNAs, whereas use of the downstream poly(A) site was reduced to 43%. There were no further downstream poly(A) sites detectable, as was the case for the wild-type GCN4 allele in the rna15 background. Whereas the GCN4mut-3 mutation alone is unable to use the wild-type poly(A) site, the additional defect in the RNA15 gene suppresses this effect and allows the use of this site again. The novel pattern of poly(A) site selection therefore reveals some of the features of each single mutant strain. Possible explanations are additional, yet unknown RNA-protein interactions between the pre-mRNA and the processing machinery.
DISCUSSION
In this report we have shown that three different positioning motifs in the GCN4 3[prime]-UTR influence poly(A) site selection in three different ways. The arrangement of these positioning motifs and their effects when point mutated are shown as a model in Figure 6. The most upstream located positioning motif is a TACA sequence which when point mutated results in completely changed positioning of the poly(A) sites. Poly(A) sites were exclusively shifted downstream of the wild-type poly(A) site. The change to the new poly(A) sites is very specific, since only one major and a minor poly(A) site nearby could be detected. In the wild-type 3[prime]-UTR these poly(A) sites are completely cryptic and were never detected. Therefore, this positioning motif seems to prevent usage of these cryptic sites. It cannot be ruled out that by introducing the point mutation a new positioning motif was created which directs the cleavage machinery to the downstream poly(A) site. However, the changes result in a more GC-rich motif, making it unlikely that it functions as a processing element, which are mostly AT-rich.
Figure 6. Model of the possible action of three different positioning motifs in the GCN4 3[prime]-UTR. The efficiency motifs (E) are underlined with black rectangles and the positioning motifs with light grey and dark grey rectangles, respectively. The wild-type poly(A) site is marked by a thick arrow. Use of the wild-type poly(A) site is enhanced by positioning motifs 1 and 3 (+). Cryptic poly(A) sites which were detected in the case of point mutated positioning motifs are indicated by dashed arrows. Since they are only used when specific positioning motifs are altered these motifs might normally prevent their use (-). Rna15p is suggested to interact with positioning motif 3 due to similarities in the poly(A) site selection in both a point mutated AAGAA motif and a wild-type GCN4 3[prime]-UTR in an rna15 mutant strain.
The second positioning motif, TAGAT, is located close to the third, the AAGAA positioning motif. In the case of the TAGAT motif the wild-type poly(A) site is still used in approximately half of all cases but an additional upstream poly(A) site turns up. Remarkably, this new poly(A) site is located directly in between the TAGAT and the AAGAA motifs. As a consequence the positioning motif AAGAA is located downstream of the poly(A) site on the 3[prime] cleavage product of the precursor mRNA. This is reminiscent of poly(A) site determinants of higher eukaryotes which include sequence motifs located upstream and downstream of the poly(A) site. The upstream motif is the highly conserved AAUAAA motif and the downstream motif is U- or GU-rich. A certain distance between these two motifs is necessary in higher eukaryotes to specify the poly(A) site (21). In yeast, the altered TAGAT motif might partially inactivate the adjacent AAGAA motif because the newly created poly(A) site concerns the first A of the AAGAA motif. These specific point mutations in the TAGAT sequence might give access to the upstream poly(A) site which was formerly hidden, e.g. by processing factors. It implies that this region must then be accessible to cleavage factors.
The role of the AAGAA motif differs from those of the other positioning motifs in the GCN4 3[prime]-UTR. Point mutations in AAGAA result in a scattered pattern of poly(A) sites. The wild-type poly(A) site was still predominantly used but several additional poly(A) sites occurred downstream of it. None of these cryptic poly(A) sites was detected in the wild-type GCN4, suggesting that an intact AAGAA motif represses their use. Additionally, AAGAA might be necessary to keep the preference for a PyA dinucleotide as selected poly(A) site. The PyA preference is well established in many yeast genes. Point mutations in PyA motifs of poly(A) sites of different genes lead to the predominant selection of other PyA sites nearby (22).
The distance between the positioning motif and poly(A) site might be an important determinant of poly(A) site selection. A computer alignment of yeast 3[prime]-UTRs revealed a significant peak of A-rich positioning motifs at a distance of ~20-25 nt from the poly(A) site (3). The AAGAA sequence of the GCN4 3[prime]-UTR exactly fits this model.
In higher eukaryotes, selection of the specific poly(A) site requires a minimal distance between the upstream and the downstream processing sequence motifs. It was suggested that binding of processing factors is the reason for the spacing requirement (21). For both the upstream and the downstream motifs, binding of specific processing factors has been demonstrated. The 160 kDa subunit of the cleavage polyadenylation specificity factor CPSF-160 binds to AAUAAA (23) and the 64 kDa subunit of the cleavage stimulation factor CstF-64 to the downstream motif. CstF-64 in vitro crosslinking experiments showed that the location of CstF-64 is important for poly(A) site selection (24).
In this report we suggest that the Rna15p processing factor, which has high similarity to CstF-64, especially in its RNA-binding domain (25), interacts with the AAGAA motif. This assumption is based on the finding that a point mutated AAGAA motif results in the same pattern of poly(A) sites as a wild-type GCN4 3[prime]-UTR in a strain carrying a temperature-sensitive allele of RNA15. In both cases, the poly(A) sites were dispersed over a broader region. A connection between the positioning motif and the Rna15p processing factor is likely; however, a direct interaction remains to be proven. SELEX experiments of Rna15p revealed a UGC/UGUAUUC/UUCC sequence as the RNA consensus motif (15). SELEX studies were also done with the homolog of Rna15p in higher eukaryotes, CstF-64. Although the RNA-binding domains are very similar, SELEX studies performed independently by two laboratories revealed different consensus sequences. In one case this was an AUGCGUUCCUCGUCC motif (26), in the other GU-rich sequences without strict consensus (15). Neither of these consensus sequences resembles the AAGAA motif. There is more similarity to the other positioning sequences of the GCN4 3[prime]-UTR. However, SELEX studies provide in vitro data which neglect the fact that processing factors tend to bind cooperatively (23).
Point mutations in the first two positioning motifs allow the specific use of other poly(A) sites in the presence of an intact AAGAA motif as putative interaction site for Rna15p. These point mutations do not change the predominance of one or a maximum of two poly(A) sites, in contrast to the multiple sites which are a result of defects in the RNA15 gene. Even at the restrictive temperature, the rna15 mutant strain showed a preference for two poly(A) sites of the GCN4mut-3 allele. In this case defective Rna15p led only to the additional use of the wild-type poly(A) site which could not be detected in a strain carrying wild-type RNA15 and the GCN4mut-3 allele. The combination of mutations in RNA15 and a sequence motif in the GCN4 3[prime]-UTR therefore results in a novel phenotype which shares some of the features of each single mutant strain. Deletion of a 15 bp spanning region of the GCN4 3[prime]-UTR performed in a former study (7) revealed that the wild-type poly(A) site and two additional downstream poly(A) sites served with equal frequency. Destruction of a whole region could therefore concern different positioning motifs and the effect need not be additive but could result in a different pattern of poly(A) site selection.
The point mutations in the GCN4 3[prime]-UTR might also have altered the RNA secondary structure, providing different access to the cleavage sites in the 3[prime]-UTR. According to an MFOLD analysis (based on 27) the GCN4 3[prime]-UTR does not fold into any stable secondary structure, which makes it unlikely that the effect of point mutations in this region would only be due to alterations of the secondary structure.
Alternatively, these sites could be targets for other proteins of the processing machinery. There are several protein factors in yeast which have been shown to possess RNA-binding activity. A possible candidate could be the Hrp1/Nab4 protein. It is part of the CFI complex, like Rna15p, and it was characterized by two different laboratories (28,29). Hrp1p/Nab4p was shown to bind (UA)n motifs (30,31). TATATA motifs mostly function as efficiency motifs of mRNA 3[prime]-end formation. The involvement of TATATA-binding factors in poly(A) site selection remains speculative, at least in those 3[prime]-UTRs in which efficiency and positioning motifs are independent of each other, like the GCN4 3[prime]-UTR. However, depletion of Hrp1p/Nab4p resulted in the use of specific upstream poly(A) sites of CYC1 and the GAL7 mRNA (29). The wild-type poly(A) site was still used but with lower frequency. The effects of a hrp1/nab4 mutation on poly(A) site selection therefore resemble the effects obtained by point mutations in the TAGAT motif upstream of the AAGAA motif. The involvement of Hrp1p/Nab4p in poly(A) site selection of the GCN4 3[prime]-UTR remains to be proven.
In summary, our data suggest more contact points between the 3[prime]-end processing machinery and the precursor mRNA in yeast than previously assumed.
ACKNOWLEDGEMENTS
We thank Sabine Eckert and Ralph Pries for critical reading and helpful discussions. This work was supported by the Deutsche Forschungsgemeinschaft, the Volkswagen-Stiftung and the Fonds der Chemischen Industrie.
REFERENCES
*To whom correspondence should be addressed. Tel: +49 551 39 3771; Fax: +49 551 39 3820; Email: gbraus{at}gwdg.de
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  X. Qu, J.-M. Perez-Canadillas, S. Agrawal, J. De Baecke, H. Cheng, G. Varani, and C. Moore The C-terminal Domains of Vertebrate CstF-64 and Its Yeast Orthologue Rna15 Form a New Structure Critical for mRNA 3'-End Processing J. Biol. Chem., January 19, 2007; 282(3): 2101 - 2115. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Zhang, E. M. Howard, and P. D. Roepe Analysis of the Antimalarial Drug Resistance Protein Pfcrt Expressed in Yeast J. Biol. Chem., December 13, 2002; 277(51): 49767 - 49775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Graber, G. D. McAllister, and T. F. Smith Probabilistic prediction of Saccharomyces cerevisiae mRNA 3'-processing sites Nucleic Acids Res., April 15, 2002; 30(8): 1851 - 1858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. T. A. Vo, M. Minet, J.-M. Schmitter, F. Lacroute, and F. Wyers Mpe1, a Zinc Knuckle Protein, Is an Essential Component of Yeast Cleavage and Polyadenylation Factor Required for the Cleavage and Polyadenylation of mRNA Mol. Cell. Biol., December 15, 2001; 21(24): 8346 - 8356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Gross and C. L. Moore Rna15 Interaction with the A-Rich Yeast Polyadenylation Signal Is an Essential Step in mRNA 3'-End Formation Mol. Cell. Biol., December 1, 2001; 21(23): 8045 - 8055. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||