Nucleic Acids Research

Strains and media

Strains are listed in Table 1. All transformations in Escherichia coli were done using DH5[alpha]. Rich yeast extract/peptone/dextrose (YPD) and synthetic complete medium lacking uracil (SCUra^-) or lacking leucine (SCLeu^-) and genetic methods were as described (26). Low adenine plates were as described elsewhere (27). Benomyl sensitivity was assayed as previously described (17). Benomyl was a gift from DuPont Co. and 5-fluoroorotic acid was purchased from US Biological (Swampscott, MA).

Table 1. Strains used in this study

Plasmids

Plasmids used in this work are listed in Table 2.

Table 2. Plasmids used in this study.

Genetic methods

The XRN1 coding sequence from nt 108 to 3894 was replaced by the LEU2 gene to give the pUC18-based plasmid pRDK226. pRDK226 was digested with HindIII and AvrII and transformed into BJ5464 to generate the xrn1::LEU2 allele. Correct integrants were identified by PCR to give RKY1806.

Overexpression of truncated Xrn1 proteins

Overnight cultures grown at 30°C in SCUra^- medium supplemented with 2% raffinose were diluted in the same medium to1.5 × 10⁷ cells/ml and galactose was added to 2%. The cultures were incubated at 30°C for 4-5 h and the cells harvested. Small scale extracts were prepared as described previously (29). Western blotting was performed as described (16), using affinity-purified polyclonal anti-Xrn1p (16) or monoclonal 12CA5 anti-HA (Berkeley Antibody Co.).

Northern blot analysis and nucleic acid substrates

Total yeast RNA was prepared and analyzed by northern blotting for ITS1 and CYH2 as described previously (17). Additional oligonucleotide probes were specific for the pre-CYH2 intron (5[prime]-CTGGACATTTTATCGAAACAAAAGAAACGTGG) and for mini-PGK1 (5[prime]-CGATAGTAATATTTATATATTTATATTTTTAAAATATTTA). The substrate for in vitro exoribonuclease assays, corresponding to the 5[prime] 261 nt of pre-CYH2 mRNA, was prepared by in vitro transcription using a T7 Ribomax kit (Promega) in the presence of [[alpha]-³²P]CTP. Specific activity of the substrate was typically 80-100 c.p.m./pmol nucleotide. The template was prepared by PCR of the endogenous yeast CHY2 gene using the primers 5[prime]-CGCGGATCCTAATACGACTCACTATAGAACAATCATCCAACTAATC, containing the T7 promoter and transcription start site for CYH2, and 5[prime]-AAGCCGTCTCAACAGTGAGATGGTA. The 5[prime]-triphosphate was hydrolyzed to 5[prime]-monophosphate using tobacco acid pyrophosphatase (Epicentre) as described elsewhere (11) and the RNA was purified by phenol extraction and ethanol precipitation. T7 [³H]DNA was prepared as described previously (31).

Single-strand DNA exonuclease assay

Reactions contained 33 mM Tris, pH 8.5, 13 mM MgCl₂, 88 µg/ml BSA, 1.8 mM dithiothreitol, heat-denatured T7 [³H]DNA (22 000 c.p.m.) and crude cell extract (0.1-1 µl) in a 30 µl volume. After incubation at 30°C for 20 min, the reactions were stopped on ice by addition of 10 µl 20 mM EDTA containing 1 mg/ml heat-denatured calf thymus DNA followed by 40 µl 1 M trichloroacetic acid. The mixture was incubated on ice for 10 min and centrifuged for 10 min and 60 µl supernatant were analyzed by liquid scintillation counting. Net activity was determined by subtracting the activity measured in a mock reaction. The limit of detection in this assay was 10% of wild-type activity.

Screen for non-complementing xrn1 missense mutations

PCR was used to generate mutations in four contiguous regions of XRN1. PCR reactions contained 100 ng plasmid template DNA, 10 pmol each primer, 5 mM MgCl₂, 0.2 mM nucleotides and 2 U Taq DNA polymerase (Promega) in a 50 µl volume. Denaturation at 94°C for 2.5 min was followed by 12 cycles of 0.5 min denaturation at 94°C, 1 min annealing at 45°C and 1 min extension at 72°C per 1 kb DNA. Ten independent PCR reactions were carried out simultaneously and were combined after PCR was complete. Region I, nominally from nt 1 to 293 of XRN1, was amplified with primers 3196 (5[prime]-TTGGGATTTTAAATGCC) and 3205 (5[prime]-AGTTTATTTTCTAAAGG) using pAJ152 as the template. The resulting PCR product (corresponding to nt -151 to 573) was co-transformed with pAJ152 digested with XhoI (nt -12) and SstI (nt 293) into AJY458 and plated on SCLeu^- low Ade plates. Region II, from nt 293 to 1335, was amplified similarly using primers 2462 (5[prime]-GAACACAGATTCCTGAG) and 2463 (5[prime]-CCAATTCTTTTTGTCC). The resulting PCR product was digested with SstI and BsmI, ligated into SstI and BsmI-digested pAJ152 and transformed into E.coli. A DNA library was prepared from the combined transformants and the library DNA was transformed into AJY458 and plated on SCLeu^- low Ade plates. Region III, nt 1335-2239, was amplified similarly, using PCR primers 3213 (5[prime]-CAGGAAAAATTATCGCC) and 2836 (5[prime]-TTTAAATGCCTCTTGG). The resulting product was digested with BsmI and SpeI and ligated into BsmI and SpeI-digested pAJ152. Region IV, nominally from nt 2239 to 3612, was amplified with primers 3216 (5[prime]-ACACAATCATTGTGTCG) and 3254 (5[prime]-GAAAGAGGCATCAAGCC). The PCR product was co-transformed with pAJ152 digested with SpeI (nt 2239) and BclI (nt 3612) into AJY458. Non-complementing XRN1 mutants were identified as non-sectoring colonies and arose at a frequency of ~2% for region I and 4% for regions II, III and IV. To eliminate nonsense mutants, 50 region I mutants, 173 region II mutants, 109 region III mutants and 96 region IV mutants were screened by western blotting for expression of Xrn1p containing the HA epitope located in the C-terminus of the protein. This yielded 10 missense mutations in region I, six in region II, one in region III and one in region IV. Region I and II mutants containing multiple mutations were sub-cloned to identify single amino acid changes responsible for the non-complementing phenotype. The region III and IV mutants each contained multiple changes and were analyzed without sub-cloning. All mutant xrn1 genes were sub-cloned to a galactose-inducible expression vector to facilitate affinity purification for in vitro analysis.

Immunoprecipitation of epitope-tagged Xrn1p

Cell cultures of BJ5464 bearing 2µ plasmids with galactose-inducible xrn1 mutants were grown in 25 ml SCLeu^- containing 2% raffinose to a density of 1.5 × 10⁷ cells/ml. For controls, cultures of BJ5464 containing wild-type HA-tagged XRN1 on pAJ251 or the empty vector, YEp351, were prepared. Galactose was added to 2% final concentration and the cells were incubated for 6 h before harvesting and storage at -80°C. Small scale extracts were prepared by glass bead disruption in 500 µl 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM DTT and 1 mM PMSF. The extracts were clarified by centrifugation at 15 000 g for 10 min at 4°C. For immunoprecipitation, 100 µg total protein (typically 10 µl extract) were diluted into 500 µl IP buffer (20 mM Tris, pH 7.5, 500 mM NaCl, 1% NP40, 0.1 mg/ml BSA, 1 mM EDTA, 1 mM PMSF, 1 µM pepstatin A and 1 µM leupeptin) on ice. A limiting amount of 12CA5 antibody (0.03-1 µg) was added and the samples were incubated on ice for 1 h. Subsequently, 15 µl of a slurry of protein A-agarose beads (Gibco) was added and the samples rocked gently at 4°C for 30 min. The beads were centrifuged, washed three times with 250 µl IP buffer and assayed immediately. Western blotting showed that under these conditions the amount of Xrn1p immunoprecipitated was proportional to the amount of 12CA5 antibody added (data not shown).

Exoribonuclease assay of immunoprecipitated wild-type and mutant Xrn1p

The protein A beads containing the bound Xrn1p-antibody complex were washed once in exonuclease buffer (20 mM Tris, pH 8.5, 75 mM NaCl, 5 mM MgCl₂, 0.1 mg/ml BSA, 0.2 mM DTT) and the beads were resuspended in 30 µl reaction mix, 20 mM Tris-HCl, pH 8.5, 75 mM NaCl₂, 5 mM MgCl₂, 0.1 mM BSA, 0.2 mM DTT and 5000 c.p.m. [³²P]pre-CYH2 RNA. Reactions were incubated for 10 min at 30°C and then stopped on ice with addition of 10 µl 1 mg/ml single-stranded DNA, 20 mM EDTA and 66 µl 0.6 M trichloroacetic acid. After 10 min on ice, the samples were centrifuged for 10 min at 4°C and 80 µl of the supernatant were removed for liquid scintillation counting. Net activity was determined by subtracting the activity measured in a mock reaction, containing an immunoprecipitate from an extract lacking HA-tagged Xrn1p. Activities were normalized to the amount of Xrn1p that was immunoprecipitated.

Figure 1. General structure of Xrn1p based on amino acid sequence alignment and protease sensitivity. Conserved domains a, b and c and domain d, unique to Xrn1p, are shaded. Percent identity for domains a and b is the percent of amino residues that are invariant among six proteins: Xrn1p and Rat1p from S.cerevisiae (4,37), ExoII and Dhp1 from S.pombe (13,18) and mXRN1p and Dhm1 from mouse (14,19). The percent identity for domain c reflects identity only among the Xrn1p sub-family, Xrn1p, Dhp1 and mXRN1p. Open arrowheads indicate sites of protease cleavage under native conditions (28). Numbers indicate the amino acid position C-terminal of cleavage. The cleavage at amino acid 1230 is approximate, based on comparisons of the relative mobilities of the proteolyzed protein and truncation proteins on SDS-polyacrylamide gels.

RESULTS

Sequence alignment of Xrn1p with related proteins suggests that the protein is composed of distinct domains (Fig. 1). The N-terminus, conserved among all family members, is acidic and contains two highly conserved regions. Previous work demonstrated a V8 protease-sensitive site at amino acid 501 (28), suggesting that the two conserved acidic blocks are separate domains. Since only these acidic domains are found in both the Rat1p and Xrn1p sub-families, they likely contain the exonuclease (see below). A middle basic domain is common to the Xrn1p sub-family. Protease digestion of Xrn1p under native conditions identified a second V8-sensitive site at amino acid 710, a chymotrypsin-sensitive site at position 723 and a trypsin-sensitive site at position 730 (28), demonstrating accessibility at the boundary of the acidic and basic domains. The basic extreme C-terminus of Xrn1p is unique to this protein and is sensitive to proteolysis. Initial purification of Xrn1p yielded a proteolyzed protein lacking the extreme C-terminus. This truncated protein retained exonuclease activity (29; unpublished data), implying that the C-terminus is a distinct domain dispensable for exonuclease activity.

A series of deletion mutants of XRN1 was constructed to identify functional domains of Xrn1p. These deletion mutants were expressed from single copy vectors for complementation assays in vivo or from high copy vectors under control of the galactose-inducible GAL10 promoter for in vitro analysis. xrn1 null mutants display a significant growth defect (4). This growth defect was complemented by mutant proteins containing small deletions from the C-terminus, mutants [Delta]1493-1528 and [Delta]1392-1528, but not by mutant proteins containing larger deletions from the C-terminus nor by proteins containing deletions of the N-terminus or internal sequences of the protein (Fig. 2 and data not shown). Consistent with this result, the two mutants [Delta]1493-1528 and [Delta]1392-1528 were also the only mutant proteins tested that complemented the inviability of an xrn1 ski2 double mutant (data not shown). All mutant proteins were present at similar high levels in total protein extracts (data not shown), suggesting that the lack of complementation was not a consequence of altered expression or protein stability.

Figure 2. Analysis of XRN1 deletion mutants. (a) Solid bars represent protein sequence remaining in the deletion mutants. Complementation of growth was scored in strain RDKY1806 containing the xrn1 alleles expressed from the single copy vector pRDK307. +, wild-type growth; -, growth like an xrn1 null mutant. High expression growth inhibition indicates inhibition of growth when expressed from a galactose-inducible promoter on a high copy vector. +++, strong growth inhibition; -, lack of growth inhibition; ND, not determined. Exonuclease activity was measured in crude extracts on ssDNA and was adjusted to account for the concentration of Xrn1p in the extracts. <10% indicates below the detection limit in this assay. (b) Strains were BJ5464 containing: XRN1 (WT) expressed from a galactose-inducible promoter on the high copy vector pRDK316; empty vector pRDK317; the indicated xrn1 mutants replacing the XRN1 gene in pRDK316. SCLeu- plates containing 2% glucose or 2% galactose were incubated at 30°C for 3 or 5 days respectively.

The extreme C-terminus of Xrn1p confers a dominant negative phenotype

Full-length wild-type Xrn1p inhibited cell growth when over-expressed from a high copy plasmid under the control of a galactose-inducible promoter (Fig. 2b). The xrn1 deletion mutants were tested for this growth inhibition phenotype. Overexpression of the mutant [Delta]1493-1528, lacking only the C-terminal 36 amino acids, conferred strong growth inhibition similar to the wild-type. Slightly larger deletions, [Delta]1392-1528, which retains a functional exonuclease, and [Delta]1206-1528, which lacks exonuclease activity, displayed modest growth inhibition. Larger deletions from the C-terminus restored growth similar to that of cells containing the empty vector, indicating a lack of growth inhibition (Fig. 2b and data not shown). The N-terminal deletions [Delta]2-397 and [Delta]1-1076 showed moderate growth inhibition, intermediate between that observed for the wild-type and mutant [Delta]1392-1528. These results indicate that the C-terminal domain that is unique to Xrn1p confers dominant negative growth inhibition when overexpressed. This activity is primarily located within amino acids 1392-1492, although amino acids from 1036 to 1391 contribute a modest level of growth inhibition. The conserved N-terminus of Xrn1p, which likely contains the exonuclease, is not necessary for the growth inhibition phenotype. Thus, these data suggest that growth inhibition is independent of nuclease function (see below).

Figure 3. Assay for complementation of xrn1 ski2 by xrn1 mutants. Strain AJY458, which contains SKI2 on a URA3 vector, was transformed with single copy vectors containing the xrn1 point mutants. Transformants were patched to a plate containing 5-fluoroorotic acid. Growth on this plate requires loss of the SKI2 URA3 plasmid and indicates complementation of xrn1 ski2.

The xrn1 deletion mutant proteins were overexpressed and assayed in vitro for exonuclease activity. We took advantage of the fact that Xrn1p demonstrates exonuclease activity on single-stranded DNA as well as RNA to avoid contaminating exoribonucleases in the crude extracts. Consistent with the in vivo complementation results, C-terminal deletions up to amino acid 1392 displayed high levels of exonuclease (60% of wild-type; Fig. 2), whereas larger deletions lacked detectable activity.

Non-complementing missense mutations in XRN1

The deletion mutant analysis did not reveal the location of the exonuclease domain of Xrn1p. Consequently, point mutations were generated and screened for non-complementing missense mutations (see Materials and Methods). PCR was used to make random mutations within four contiguous regions of the gene corresponding to amino acids 1-1203. Nineteen non-complementing xrn1 missense mutants were identified based on their inability to complement an xrn1 ski2 double mutant. Seventeen of the mutants were within regions I (nt 1-293) and II (nt 294-1335). Of these, seven contained two nucleotide changes and one contained three nucleotide changes. Only two of the mutants containing multiple nucleotide changes resulted in second amino acid changes, E176G F29L and E178G Q401L. These two mutants were sub-cloned to separate the mutations. The single mutations E176G and E178G both conferred a null phenotype, whereas F29L and Q401L showed no obvious phenotype. Of the 17 region I and II mutants, three were represented twice, giving 14 different single amino acid changes of highly conserved residues mapping from amino acid 37 to 201 (Figs 3 and 4 and Table 3). Two additional mutants, referred to as III-17, from region III (nt 1336-2239), and IV-74, from region IV (nt 2240-3612), contained multiple amino acid changes between amino acids 592 and1197 and these mutants were analyzed without sub-cloning. Mutant Q97R showed a low level of complementation of xrn1 ski2 (Fig 3), consistent with having the highest level of exoribonuclease activity among the mutants (Table 3). Mutant G87D showed very weak complementation (Fig. 3) but no measurable exonuclease activity in vivo (Figs 5 and 6) or in vitro (Table 3). Further characterization of this mutant is required to understand how it provides complementing activity. It is possible that the mutant protein acts by efficiently sequestering substrates without degradation. When the xrn1 point mutants were tested for their ability to complement the synthetic lethality of xrn1 ski3 double mutants, results similar to those shown in Figure 3 were obtained (data not shown), suggesting that the exonuclease function of Xrn1p is required in ski3 mutants as well.

Table 3.
^aActivity is expressed relative to that of wild-type HA-tagged Xrn1p, after correcting for activity in reactions lacking HA-tagged Xrn1p.

Figure 4. Homology alignment and positions of amino acid changes identified in xrn1 mutants. The top three lines of alignment are the Xrn1p sub-family: ExoII, from S.pombe; Dhm2, from mouse. The lower three lines are the Rat1p sub-family: Dhp1, from S.pombe; Dhm1 from mouse. Arrows indicate residues affecting exonuclease activity that were altered in xrn1 point mutants. Asterisks represent residues that could be changed in Xrn1p (F29L) or Rat1p (M36T) without affecting nuclease activity. The open circle indicates the position of the amino acid change (P90L) in the xrn1-10^ts allele.

Figure 5. Northern blot analysis of in vivo substrates of Xrn1p. Ten micrograms of total RNA were separated on a 1.2% agarose-formaldehyde gel, transferred to nylon membrane and probed with a probe specific for ITS1 or CYH2. Strains were RDKY1977 containing the indicated xrn1 alleles in the single copy vector pAJ152 and were grown in SCLeu^-. III-17 and IV-74 contain multiple amino acid changes (see Table 3).

Figure 6. Northern blot analysis of nonsense codon-containing mRNAs. Conditions were as described in Figure 5 except that the strains used here also contained the mini-PGK1 allele on pRIPGKH2(3)[Delta]1IN1 (30) and strains were grown in SCLeu^-Ura^-. The blot was probed simultaneously with an intron-specific probe for pre-CYH2 and with a probe specific for mini-PGK1.

Northern blot analysis of xrn1 mutants

Northern blotting was done on total RNA isolated from xrn1 mutant strains containing the xrn1 point mutants expressed from single copy vectors. Blots were probed for ITS1, CYH2 and mini-PGK1. ITS1 is a product of rRNA processing that is rapidly degraded in wild-type cells by Xrn1p (8). As seen in Figure 5, all of the mutants tested displayed stabilization of ITS1 in vivo, indicating that all were defective for exonuclease activity. The degree of stabilization and the nature of the ITS1 signal varied among the mutants. Whereas most mutants stabilized intact ITS1, some of the mutants, e.g. Q97E and E178D, showed broadening of the ITS1 band to lower molecular weight, suggesting partial degradation.

Analysis of pre-CYH2 RNA also demonstrated defects in in vivo degradation, but in addition revealed novel degradation intermediates. CYH2 pre-mRNA is inefficiently spliced in yeast and the pre-mRNA is rapidly degraded in the cytoplasm by Xrn1p after decapping by the deadenylation-independent nonsense-mediated decay pathway (3). Again, varying degrees of stabilization of pre-CYH2 were observed. However, some of the mutants, notably N37D, H41R, Q97E, Q97R and E178D, gave rise to new bands corresponding to CYH2 pre-mRNA ~100 nt shorter than full length. These new species represented RNAs that had been only partially degraded from the 5[prime]-ends and were not due to altered processing from the 3[prime]-end, since they contained poly(A) tails which could be removed by RNase H digestion in the presence of oligo(dT) without altering the relative position of the pre-CYH2 to the degradation intermediates (data not shown). Pre-CYH2 RNA was also analyzed with an intron-specific probe (Fig. 6). Simultaneously, these RNAs were probed with a mini-PGK1-specific probe. Both of these RNAs are degraded by nonsense-mediated decay. In both cases, discrete degradation intermediates were observed. Interestingly, amino acid substitution of H41 and E178 gave rise to different phenotypes dependent on the amino acid substitution. Conservative changes that maintained charge (H41R and E178D) resulted in mutants that showed degradation intermediates, whereas the non-conservative changes H41D and E178G resulted in a complete blockage of degradation, indicated by stabilization of the full-length RNAs. Some transcript-specific differences were observed. For example, mutant C201Y displayed wild-type degradation of mini-PGK1 mRNA but accumulated degradation intermediates of CYH2 pre-mRNA.

The degradation intermediates resulted from mutant Xrn1p processing, since they were not seen in the absence of Xrn1p (Figs 5b and 6). Also, accumulation of these intermediates was not observed when the mutant protein N37D or E178D was over-expressed in a wild-type strain, indicating that wild-type Xrn1p can degrade these RNAs in the presence of mutant Xrn1p (data not shown). This suggests that the mutant proteins do not remain tightly bound to the substrate, impeding degradation by wild-type protein. Taken together, the novel degradation intermediates appear to result from mutant Xrn1p being blocked either by proteins bound to the transcript or by secondary structure in the transcript.

In vitro exonuclease activities

The mutant and wild-type Xrn1p proteins were overexpressed and purified from crude extracts of wild-type yeast by immunoprecipitation, taking advantage of the HA epitope in their C-terminus. All mutant and wild-type proteins were expressed at similar levels, suggesting that all proteins had similar stability. The immunoprecipitated proteins were then assayed for exoribo-nuclease activity without further purification. For these assays, the substrate was an in vitro synthesized transcript corresponding to the first 261 nt of the 5[prime]-end of CYH2 pre-mRNA. Xrn1p concentration in the assays was controlled by titrating the Xrn1p proteins with limiting amounts of anti-HA antibody. Under these conditions, the amount of Xrn1p in the assays was proportional to the amount of anti-HA used for immunoprecipitation (data not shown). Results of these assays are shown in Table 3. The in vitro activity closely paralleled the in vivo activity observed on ITS1, pre-CYH2 and mini-PGK1. These assays were remarkably free of contaminating exoribonucleases and the level of detection was <2% of wild-type Xrn1p activity.

Figure 7. Benomyl hypersensitivity of xrn1 exonuclease mutants. Cultures of strain RDKY1977, containing the indicated xrn1 alleles on the single copy vector pAJ152, were spotted onto YPD plates containing 0 or 15 µg/ml benomyl. Plates were incubated at 30°C for 2 days.

The mutant proteins were also assayed for single-strand DNA exonuclease activity. This assay was done with crude extracts without immunopurification. Under these conditions, overexpressed Xrn1p accounts for ~90% of the total exonuclease activity in these extracts. As seen in Table 3, all mutant proteins displayed reduced DNA exonuclease activity that in general corresponded to the decrease in exoribonuclease activity, suggesting that these activities reside in the same active site. One exception, Q97R, displayed much lower relative DNase activity compared with RNase activity. It is not clear at present what accounts for this difference, but it may be due to differences in binding deoxyribose compared with ribose. The correlation of RNase and DNase activities suggests that these activities are both manifest by the same active site.

xrn1 nuclease-deficient mutants do not complement the benomyl-hypersensitive phenotype of xrn1 null mutants

Benomyl destabilizes microtubules and mutants that display microtubule defects often confer hypersensitivity to benomyl (32). Such hypersensitivity of xrn1 mutants has led to the suggestion that Xrn1p may interact directly with tubulin (22). We tested the various xrn1 point mutants for their ability to complement the benomyl sensitivity of an xrn1 null mutant. There was a general correlation between the severity of the exonuclease defect and inability to complement benomyl hypersensitivity. Mutants that lacked detectable exonuclease activity, e.g. E178G and H41D, were at least as sensitive as an xrn1 null mutant, whereas Q97R, which displayed 60% of wild-type exonuclease activity and partially restored growth to an xrn1 ski2 double mutant, partially restored wild-type sensitivity to benomyl (Fig. 7).

Growth inhibition from overexpression of Xrn1p does not depend on exonuclease

The dominant negative phenotype that results from overexpression of the C-terminus of Xrn1p was also observed for all of the missense mutants (Fig. 8). Thus, growth inhibition is not due to overexpression of the nuclease activity of Xrn1p, but rather some other as yet unidentified function of the protein.

Figure 8. Growth inhibition from overexpression of XRN1 does not require exonuclease activity. Cultures of strain BJ5464, containing the indicated xrn1 alleles under control of a galactose-inducible promoter on the high copy plasmid pAJ251, were patched onto SCLeu^- plates containing 2% glucose or 2% galactose and incubated at 30°C for 3 days.

DISCUSSION

Xrn1p is the founding member of a family of eukaryotic exoribonucleases. It is the major activity in the cytoplasm of yeast responsible for mRNA degradation and is also responsible for the elimination of rRNA processing fragments and improperly processed mRNAs subject to nonsense-mediated decay. In yeast, a second but essential 5[prime]-exoribonuclease, encoded by RAT1, is found in the nucleus (17). These two enzymes are functionally interchangeable, although normally restricted to the cytoplasm and nucleus respectively (17). Homologs of Xrn1p and Rat1p have been identified from higher eukaryotes, includingD.melanogaster (D.Till, B.Linz, J.Seago, J.McClellan, J.McCarthy and S.F.Newbury, personal communication) and mouse (14,19), however, no homologs of these proteins have been identified from prokaryotic cells.

Xrn1p and its homologs are large proteins that contain exonuclease as well as nucleic acid binding activities, although they share no obvious sequence motifs with other classes of proteins. In addition, Xrn1p has been implicated in additional cellular processes, including interaction with tubulin (22). Thus, it is of interest to identify the structural motifs of these proteins responsible for their biochemical properties and to identify mutations within these motifs that eliminate specific biochemical functions to correlate these functions with observed phenotypes.

This family of proteins can be aligned based on primary amino acid sequence, revealing an acidic N-terminal half of the protein composed of two domains of striking homology common to all family members (Fig. 1). Since this acidic region of the protein is the only portion common to this family of exonucleases, it likely contains the exonuclease. C-terminal to this, the Xrn1p and Rat1p sub-families diverge in amino acid sequence. However, the sub-families each contain basic domains and thus may have conservation at a gross structural level. The extreme C-terminus of Xrn1p, again basic, is unique to this protein. The apparent arrangement of these proteins into domains based on sequence alignments is reflected in the susceptibility of Xrn1p to proteolysis under native conditions. Protease cleavage of Xrn1p for amino acid sequencing using several different proteases yielded cleavage at amino acid 501, between the two blocks of amino acid conservation within the N-terminal acid half of the protein and after amino acids 710, 723 and 730, suggesting that the boundary between the highly conserved acidic N-terminus and the following basic domain is readily accessible to proteases (28). In addition, the extreme C-terminus, unique to Xrn1p, is highly susceptible to proteolysis during purification (29).

Surprisingly, although the protein appears to be a linear assemblage of domains, deletion of any portion of the homologous domains rendered the protein completely inactive in vivo and in vitro, even though subsequent point mutation analysis suggested that the N-terminus contains the exonuclease. Thus, this analysis failed to localize the exonuclease domain of the protein. We do not yet understand why this protein is not amenable to deletions, but it may reflect changes in the overall net charge of the protein when portions of an acidic or basic domain are deleted.

Deletions of the extreme C-terminus do not affect most in vivo functions, but the C-terminus does confer growth inhibition when the protein is overexpressed. This growth inhibition depended primarily on amino acids 1392-1492 and was observed for constructs containing large N-terminal deletions. However, high expression growth inhibition was observed for exonuclease-deficient point mutants, demonstrating that it was not due to the overexpression of the exonuclease activity. It is possible that the C-terminus interacts with other essential cellular components, titrating them to below critical levels. Two-hybrid analysis with the C-terminal domain should identify such factors.

Point mutations identify residues critical for exonuclease function

The screen employed for identifying point mutants required that such mutants not complement an xrn1 ski2 double mutant. All xrn1 mutations identified in this screen reduced or abolished detectable exonuclease activity, suggesting that synthetic lethality with a ski2 mutation (23) is due specifically to a lack of Xrn1p exonuclease activity. This was not a foregone conclusion. Xrn1p is also an RNA binding protein and defects in RNA binding could conceivably be lethal in combination with a ski2 mutation. In addition, Ski2p suppresses translation of poly(A)^- mRNA (24,33). Ski2p may be necessary to confer specificity of ribosomes for poly(A) mRNA or by acting in a degradation pathway that degrades such mRNA. The requirement for nuclease activity for complementation of a ski2 mutation is consistent with the finding that SKI2 is necessary for a 3[prime] mRNA degradation pathway (25). In this model, residual 3[prime] degradation controlled by Ski2p is essential in the absence of the major 5[prime] degradation pathway, which requires Xrn1p. This screen did not identify mutations in other regions of the protein or shed light on other functions of Xrn1p. To identify other functions of Xrn1p, we are currently screening additional mutants that are synthetically lethal with an xrn1 null mutant for complementation by an xrn1 exonuclease-deficient mutant.

The Xrn1p family of proteins does not contain amino acid sequences that are readily identifiable as exonuclease or RNA binding motifs. However, all single point mutants identified in this screen reduced or abolished detectable nuclease activity and were localized to the highly conserved N-terminal domain of Xrn1p. At least some of these residues are likely within the active site of the exonuclease. Like many nucleases, Xrn1p requires Mg²⁺ for activity. Structural and biochemical studies of other nucleases have shown that such divalent cations are typically coordinated by carboxylates (34-36). In Xrn1p, several invariant carboxylate residues, D86, E176 and E178, appeared to be essential for function. Replacement of E178 with aspartate, a conservative change that maintains a carboxylate moiety, reduced the nuclease activity but did not eliminate it, whereas the more drastic change to a glycine eliminated detectable activity altogether. Possibly E178 coordinates a metal ion and replacing E178 with aspartate leads to a less favorable coordination. Similarly, H41 is important for nuclease function. Changing H41 to arginine maintains the charge and results in a protein with reduced activity, whereas the more drastic change to aspartate eliminates detectable activity.

During the course of this work, the xrn1-10 temperature-sensitive mutation (23) was identified as P90L. This mutation disrupts exonuclease activity at non-permissive temperature (A.Johnson, unpublished results) and lies within a highly conserved block of amino acids. In addition, changing F29 in Xrn1p or changing M36 in Rat1p had no significant affect on enzyme activity (A.Johnson, unpublished results). Thus, some changes of highly conserved residues do not significantly affect function, suggesting that the mutations identified in this screen were particularly critical for nuclease activity.

All mutants were defective to varying degrees for exonuclease activity measured in vivo by northern blot analysis of several different RNAs and by in vitro assays. In general, the in vivo activities correlated with the in vitro activities. In addition, there was a good correlation between the amount of residual exonuclease activity and the ability to complement an xrn1 ski2 double mutant. Furthermore, mutations that affected the exoribonuclease activity of Xrn1p also affected in a similar fashion its activity on single-stranded DNA, suggesting that these two activities are manifested by the same active site.

An interesting class of xrn1 mutants yielded novel degradation intermediates on mRNAs targeted for rapid degradation in the deadenylation-independent nonsense-mediated decay pathway. These mutants displayed reduced exonuclease activity measured in vitro and were unable to completely degrade certain RNAs in vivo. It is possible that these mutants have reduced processivity, although the exact nature of their defect will require more extensive in vitro characterization with purified mutant proteins. These mutants are not dominant, suggesting that the novel RNA intermediates are not due to mutant Xrn1p binding tightly to RNA and thereby inhibiting the progress of wild-type Xrn1p. The intermediates most likely are due to blockage by either the intrinsic structure of the RNA or proteins bound to the RNA. We are currently examining the nature of the block on the RNA. Preliminary in vitro results suggest that although the mutant proteins have reduced exonuclease activity, they are not blocked at a discrete point on pre-CYH2, suggesting that the degradation block is formed in vivo and may be due to proteins on the RNA. Possibly, Xrn1p is responsible for displacing proteins on the transcript, for example ribosomes, as the RNA is degraded.

It has been reported that Xrn1p is a tubulin binding protein based on genetic and in vitro studies (22), which concur with the initial report that xrn1 mutants are hypersensitive to the tubulin destabilizing drug benomyl (4). We tested the xrn1 point mutants described here for their ability to complement benomyl hypersensitivity. There was a good correlation between the ability to complement and the amount of residual exonuclease activity displayed by the mutant. Mutants with no detectable exonuclease activity were unable to complement the hypersensitivity. Protein-protein interaction between Xrn1p and tubulin would most likely depend on surface contacts and not depend on the enzymatic activity of Xrn1p. Since at least some of the xrn1 point mutants described here likely alter residues within the active site of the enzyme and may not perturb other activities or structure of the protein, the lack of complementation strongly suggests that the putative interaction with tubulin is indirect. That these mutants do not greatly perturb protein structure is supported by the fact that they still display growth inhibition when overexpressed. An indirect connection between Xrn1p and tubulin is also suggested by the finding that mislocalization of a second exoribonuclease, Rat1p, from the nucleus to the cytoplasm complements benomyl hypersensitivity (17).

ACKNOWLEDGEMENTS

We would like to thank M.Burkart and Y.Xue for assistance in plasmid construction and S.Ross for help with the point mutation screen. We thank P.Ljundahl for providing a series of xrn1 deletions prepared for sequence analysis, S.Peltz for providing plasmid DNA and DuPont for kindly providing benomyl. This work was supported in part by NIH fellowship grant GM13594 to A.J., NIH grant GM53655 to A.J. and NIH grant GM29383 to R.K.

REFERENCES

1. Larimer,F.W., Hsu,C.L., Maupin,M.K. and Stevens,A. (1992) Gene, 120, 51-57. MEDLINE Abstract

2. Decker,C.J. and Parker,R. (1994) Trends Biochem. Sci., 19, 336-340. MEDLINE Abstract

3. Muhlrad,D. and Parker,R. (1994) Nature, 370, 578-581. MEDLINE Abstract

4. Kim,J., Ljungdahl,P.O. and Fink,G.R. (1990) Genetics, 126, 799-812. MEDLINE Abstract

5. Hsu,C.L. and Stevens,A. (1993) Mol. Cell. Biol., 13, 4826-4835. MEDLINE Abstract

6. Muhlrad,D., Decker,C.J. and Parker,R. (1994) Genes Dev., 8, 855-866. MEDLINE Abstract

7. Muhlrad,D., Decker,C.J. and Parker,R. (1995) Mol. Cell. Biol., 15, 2145-2156. MEDLINE Abstract

8. Stevens,A., Hsu,C.L., Isham,K.R. and Larimer,F.W. (1991) J. Bacteriol., 173, 7024-7028. MEDLINE Abstract

9. Stevens,A. (1980) J. Biol. Chem., 255, 3080-3085. MEDLINE Abstract

10. Decker,C.J. and Parker,R. (1993) Genes Dev., 7, 1632-1643. MEDLINE Abstract

11. Poole,T.L. and Stevens,A. (1997) Biochem. Biophys. Res. Commun., 235, 799-805. MEDLINE Abstract

12. Liu,Z. and Gilbert,W. (1994) Cell, 77, 1083-1092. MEDLINE Abstract

13. Szankasi,P. and Smith,G.R. (1996) Curr. Genet., 30, 284-293. MEDLINE Abstract

14. Bashkirov,V.I., Scherthan,H., Solinger,J.A., Buerstedde,J.M. and Heyer,W.D. (1997) J. Cell Biol., 136, 761-773. MEDLINE Abstract

15. Shobuike,T., Sugano,S., Yamashita,T. and Ikeda,H. (1997) Gene, 191, 161-166. MEDLINE Abstract

16. Heyer,W.-D., Johnson,A.W., Reinhart,U. and Kolodner,R.D. (1995) Mol. Cell. Biol., 15, 2728-2736. MEDLINE Abstract

17. Johnson,A.W. (1997) Mol. Cell. Biol., 17, 6122-6130. MEDLINE Abstract

18. Sugano,S., Shobuike,T., Takeda,T., Sugino,A. and Ikeda,H. (1994) Mol. Gen. Genet., 243, 1-8. MEDLINE Abstract

19. Shobuike,T., Sugano,S., Yamashita,T. and Ikeda,H. (1995) Nucleic Acids Res., 23, 357-361. MEDLINE Abstract

20. Kipling,D., Tambini,C. and Kearsey,S.E. (1991) Nucleic Acids Res., 19, 1385-1391. MEDLINE Abstract

21. Tishkoff,D.X., Rockmill,B., Roeder,G.S. and Kolodner,R.D. (1995) Genetics, 139, 495-509. MEDLINE Abstract

22. Interthal,H., Bellocq,C., Bähler,J., Bashkirov,V.I., Edelstein,S. and Heyer,W.-D. (1995) EMBO J., 14, 1057-1066. MEDLINE Abstract

23. Johnson,A.W. and Kolodner,R.D. (1995) Mol. Cell. Biol., 15, 2719-2727. MEDLINE Abstract

24. Masison,D.C., Blanc,A., Ribas,J.C., Carroll,K., Sonenberg,N. and Wickner,R.B. (1995) Mol. Cell. Biol., 15, 2763-2771. MEDLINE Abstract

25. Anderson,J. and Parker,R. (1998) EMBO J., 17, 1497-1506. MEDLINE Abstract

26. Rose,M.D., Winston,F. and Hieter,P. (1990) Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

27. Kranz,J.E. and Holm,C. (1990) Proc. Natl. Acad. Sci. USA, 87, 6629-6633. MEDLINE Abstract

28. Tishkoff,D.X., Johnson,A.W. and Kolodner,R.D. (1991) Mol. Cell. Biol., 11, 2593-2608. MEDLINE Abstract

29. Johnson,A.W. and Kolodner,R.D. (1991) J. Biol. Chem., 266, 14046-14054. MEDLINE Abstract

30. Peltz,S.W., Brown,A.H. and Jacobson,A. (1993) Genes Dev., 7, 1737-1754. MEDLINE Abstract

31. Richardson,C.C. (1966) J. Mol. Biol., 15, 49-61. MEDLINE Abstract

32. Schatz,P.J., Solomon,F. and Botstein,D. (1988) Mol. Cell. Biol., 6, 3722-3733.

33. Bernard,L., Carroll,K., Valle,R.C.P. and Wickner,R.B. (1998) Mol. Cell. Biol., 18, 2688-2696.

34. Beese,L.S. and Steitz,T.A. (1991) EMBO J., 10, 25-33. MEDLINE Abstract

35. Mueser,T.C., Nossal,N.G. and Hyde,C.C. (1996) Cell, 85, 1101-1112. MEDLINE Abstract

36. Xu,Y., Derbyshire,V., Ng,K., Sun,X.C., Grindley,N.D.F. and Joyce,C.M. (1997) J. Mol. Biol., 268, 284-302.

37. Amberg,D.C., Goldstein,L.A. and Cole,C.N. (1992) Genes Dev., 6, 1173-1189. MEDLINE Abstract

---

*To whom correspondence should be addressed at: Department of Microbiology, University of Texas at Austin, Austin, TX 78712-1095, USA.Tel: +1 512 475 6350; Fax: +1 512 471 7088; Email: arlen@mail.utexas.edu
⁺Present address: Center for Molecular Medicine and Therapeutics, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
^§Present address: Ludwig Institute for Cancer Research, University of California San Diego School of Medicine, La Jolla, CA 92093-0660, USA

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 30 Jul 1998
Copyright©Oxford University Press, 1998.
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				S. Kramer, R. Queiroz, L. Ellis, H. Webb, J. D. Hoheisel, C. Clayton, and M. Carrington Heat shock causes a decrease in polysomes and the appearance of stress granules in trypanosomes independently of eIF2{alpha} phosphorylation at Thr169 J. Cell Sci., September 15, 2008; 121(18): 3002 - 3014. [Abstract] [Full Text] [PDF]

				N.-J. Hung, K.-Y. Lo, S. S. Patel, K. Helmke, and A. W. Johnson Arx1 Is a Nuclear Export Receptor for the 60S Ribosomal Subunit in Yeast Mol. Biol. Cell, February 1, 2008; 19(2): 735 - 744. [Abstract] [Full Text] [PDF]

				C.-H. Li, H. Irmer, D. Gudjonsdottir-Planck, S. Freese, H. Salm, S. Haile, A. M. Estevez, and C. Clayton Roles of a Trypanosoma brucei 5'->3' exoribonuclease homolog in mRNA degradation RNA, December 1, 2006; 12(12): 2171 - 2186. [Abstract] [Full Text] [PDF]

				J. Hedges, Y.-I Chen, M. West, C. Bussiere, and A. W. Johnson Mapping the Functional Domains of Yeast NMD3, the Nuclear Export Adapter for the 60 S Ribosomal Subunit J. Biol. Chem., December 1, 2006; 281(48): 36579 - 36587. [Abstract] [Full Text] [PDF]

				L. WANG, M. S. LEWIS, and A. W. JOHNSON Domain interactions within the Ski2/3/8 complex and between the Ski complex and Ski7p RNA, August 1, 2005; 11(8): 1291 - 1302. [Abstract] [Full Text] [PDF]

				C. T. Brew and T. C. Huffaker The Yeast Ubiquitin Protease, Ubp3p, Promotes Protein Stability Genetics, November 1, 2002; 162(3): 1079 - 1089. [Abstract] [Full Text] [PDF]

				Y. Zuo and M. P. Deutscher Exoribonuclease superfamilies: structural analysis and phylogenetic distribution Nucleic Acids Res., March 1, 2001; 29(5): 1017 - 1026. [Abstract] [Full Text] [PDF]

				Y. Xue, X. Bai, I. Lee, G. Kallstrom, J. Ho, J. Brown, A. Stevens, and A. W. Johnson Saccharomyces cerevisiae RAI1 (YGL246c) Is Homologous to Human DOM3Z and Encodes a Protein That Binds the Nuclear Exoribonuclease Rat1p Mol. Cell. Biol., June 1, 2000; 20(11): 4006 - 4015. [Abstract] [Full Text]

				J. A. Solinger, D. Pascolini, and W.-D. Heyer Active-Site Mutations in the Xrn1p Exoribonuclease of Saccharomyces cerevisiae Reveal a Specific Role in Meiosis Mol. Cell. Biol., September 1, 1999; 19(9): 5930 - 5942. [Abstract] [Full Text] [PDF]

				J. H.-N. Ho and A. W. Johnson NMD3 Encodes an Essential Cytoplasmic Protein Required for Stable 60S Ribosomal Subunits in Saccharomyces cerevisiae Mol. Cell. Biol., March 1, 1999; 19(3): 2389 - 2399. [Abstract] [Full Text] [PDF]

This Article Abstract Print PDF (448K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (15) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Page, A. M. Articles by Johnson, A. W. Search for Related Content PubMed PubMed Citation Articles by Page, A. M. Articles by Johnson, A. W. Social Bookmarking          What's this?