Rea1, a Dynein-related Nuclear AAA-ATPase, Is Involved in Late rRNA Processing and Nuclear Export of 60 S Subunits*

Rea1, the largest predicted protein in the yeast genome, is a member of the AAA+ family of ATPases and is associated with pre-60 S ribosomes. Here we report that Rea1 is required for maturation and nuclear export of the pre-60 S subunit. Rea1 exhibits a predominantly nucleoplasmic localization and is present in a late pre-60 S particle together with members of the Rix1 complex. To study the role of Rea1 in ribosome biogenesis, we generated a repressible GAL::REA1 strain and temperature-sensitive rea1 alleles. In vivo depletion of Rea1 results in the significant reduction of mature 60 S subunits concomitant with defects in pre-rRNA processing and late pre-60 S ribosome stability following ITS2 cleavage and prior to the generation of mature 5.8 S rRNA. Strains depleted of the components of the Rix1 complex (Rix1, Ipi1, and Ipi3) showed similar defects. Using an in vivo 60 S subunit export assay, a strong accumulation of the large subunit reporter Rpl25-GFP (green fluorescent protein) in the nucleus and at the nuclear periphery was seen in rea1 mutants at restrictive conditions.

The synthesis of ribosomes is one of the major and most energy-consuming processes in the cell. In Saccharomyces cerevisiae, ribosome biogenesis begins in the nucleolus with the transcription of two rRNA precursors, the 35 S and the pre-5 S RNA, by RNA polymerases I and III, respectively. The 35 S pre-rRNA contains the sequences for the mature 18 S, 5.8 S, and 25 S rRNAs, two external transcribed spacers (ETS) 1 and two internal transcribed spacers (ITS). During the maturation process, the pre-rRNA has to undergo a number of modifications and is subjected to cleavages and trimming events. At least 170 accessory proteins including putative RNA helicases, endo-and exonucleases, and putative GTPases and AAA-ATPases as well as small nucleolar ribonucleoprotein particles are involved in the maturation of rRNA and its assembly into ribosomal subunits (1,2).
Concomitant with rRNA processing, ribosomal and non-ribosomal proteins are assembled on the pre-35 S rRNA, giving rise to a large 90 S pre-ribosomal particle (see Fig. 6B) (3,4). The initial cleavages at sites A 0 -A 2 separate the two subunits. The pre-40 S subunit is exported relatively rapidly to the cytoplasm, where it undergoes further processing. In contrast, the maturation of the large subunit continues in the nucleoplasm with recruitment of 60 S-specific biogenesis factors and further processing of the 27 S pre-rRNA. This includes the late cleavage and processing in the ITS2 region, which generates mature 5.8 S and 25 S rRNA.
In the last few years, the maturation of 40 S and 60 S pre-ribosomes has been extensively analyzed by purification of pre-ribosomal particles (5)(6)(7)(8)(9)(10). Interestingly, a large number of non-ribosomal proteins were identified in pre-60 S particles without an assigned function in RNA metabolism. In contrast to the pre-40 S particles, the nascent 60 S particles contain several putative GTPases and AAA-type ATPases (2,11).
To understand the events of ribosome biogenesis, we previously purified pre-ribosomal 60 S particles, which represent different maturation states from early nucleolar through cytoplasmic export (5). One of these, the late nucleoplasmic Rix1 particle, was selected for further study. The Rix1 particle is specifically enriched in the products of three uncharacterized open reading frames, YHR085w (Ipi1), YNL182c (Ipi3), and Rea1. Ipi1 has a single ARM repeat motif (12), whereas Ipi3 contains two WD40 domains. In contrast, Rea1, which at 560 kDa is the largest protein identified in the yeast genome, contains several interesting domains and homologies. The N terminus of Rea1 possesses six AAA ATPase protomers and a C-terminal region containing the MIDAS (metal ion-dependant adhesion site) motif. Furthermore, sequence analysis has indicated relatedness to dynein (13).
In this study, we characterize Rea1 and the Rix1 complexcontaining pre-60 S particle. Our results show that Rea1 and the Rix1 complex exhibit similar subcellular localization and a similar late ITS2 rRNA processing defect. Although the Rix1 complex mutants accumulate the Rpl25-GFP reporter throughout the nucleoplasm, rea1 temperature-sensitive (ts) mutants also show a later defect, with accumulation around the nuclear periphery. Our data demonstrate that Rea1 and the Rix1 complex play an essential role in maturation and nuclear export of nascent 60 S subunits from the nucleoplasm to the cytoplasm.

EXPERIMENTAL PROCEDURES
Yeast Strains and Plasmids-Genomic integration of GFP (HIS3MX6 marker) as a C-terminal tag into yeast strains to create fusion proteins of Ipi1 (strain DS1-2b, MATa, ura3, trp1, his3, leu2) and Ipi3 (strain JBa, MATa trp1 ura3 ade2 ade3 leu2 his3) was performed as described (14). For construction of GAL1::GFP-REA1 and * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Generation of Rea1 ts Alleles-20 g of pLEU2-REA1 plasmid DNA was incubated in 500 l of 2 M hydroxylamine buffer for 20 h at 55°C (20). The REA1 shuffle strain was transformed with the mutagenized plasmid, and cells were streaked on plates containing selection medium lacking leucine for 5 days at 23°C in the dark. About 2000 single colonies were picked and streaked twice on 5-fluoroorotic acid-containing plates at 23°C before it was streaked on YPD plates at 23 and 37°C. Two ts rea1 mutants (rea1-7 and rea1- 21) were derived from this screen from which the mutagenized pLEU2-REA1 plasmids were recovered and reintroduced into the REA1 shuffle strain to verify the ts phenotype. The ts mutant rea1-21 was used for further characterization.
Affinity Purification and MgCl 2 Treatments-Affinity purification of TAP-tagged proteins was performed as described previously using 2-6 liters of yeast culture (21). Normal buffer composition for purification was 0.1 M NaCl, 50 mM Tris, pH 7.5, 1.5 mM MgCl 2 , 0.75% Nonidet P-40. For MgCl 2 salt elution, the Rix1-TAP purification was as described until after the beads were bound to calmodulin (CaM)-Sepharose. The beads were then incubated for 10 min at 30°C with LB-buffer plus 100 mM MgCl 2 , the eluate was collected, and the column was washed with 5 ml of LB ϩ 100 mM MgCl 2 . After repeating the procedure with LB ϩ 200 mM MgCl 2 , the sample was eluted from the beads by boiling for 10 min in SDS sample buffer (0.5% SDS, 50 mM NaCl, and 10 mM Tris, pH 7.5).
RNA Analysis-Northern hybridization and primer extension were performed as described on whole cell extracts (22,23). Oligonucleotides Miscellaneous-Western blot analyses were performed according to Ref. 24. Fluorescence microscopy was done as described (11). The fluorescence-based visual assay to analyze the nuclear export of large and small ribosomal subunits using the Rpl25-GFP and Rps2-GFP reporters, respectively, in living cells was performed according to Refs. 18, 19, and 25. Sedimentation analysis of ribosomes under low salt conditions by sucrose gradient centrifugation was performed as described (10). Mass spectrometry using tryptic digests from Coomassie Blue-stained bands derived from SDS-PAGE was performed as described (10). Synthetic lethality was determined by tetrad dissection of either the rix1-1 (10) or the rix7-1 (11) strain mated to the rea1 deletion strain complemented with the pRS316-REA1 plasmid. Haploid progeny containing either the rix1-1 or the rix7-1 mutation and the rea1 deletion were transformed with pRS315 plasmids containing no insert or rea1-7 or rea1-21 mutant alleles. Strains were considered synthetic lethal if they could not grow on 5-fluoroorotic acid, i.e. if they could not lose the wild-type pRS316-REA1 plasmid.

High Salt Dissociates Rea1 and 60 S Subunits from the Rix1
Complex-We have previously reported the isolation and identification of 60 S pre-ribosomal particles, which are characterized by the presence of different ribosomal precursors depending on the maturation state of each of the particles (5, 10). The Rix1 particle represents a late intermediate in 60 S biogenesis and is highly enriched in the AAA-ATPase Rea1/Mdn1, as well as in two other non-ribosomal proteins, Yhr085/Ipi1 and Ynl182/Ipi3 ( Fig. 1A; see also Ref. 5). All three proteins are specifically associated with the Rix1 particle since they are largely absent both from earlier and from later pre-60 S ribosomes. The observation that Rix1, Ipi1, Ipi3, and Rea1 coenrich during biochemical purification indicates that these four proteins could be organized in a complex attached to nascent 60 S subunits. To test this possibility, we affinity-purified Rix1-TAP by the tandem affinity purification method. However, instead of eluting the purified complexes from the final CaM-Sepharose with EGTA (as in Fig. 1A), we treated the beads with increasing concentrations of MgCl 2 (Fig. 1B). The additional salt should release the interacting proteins that are not tightly bound with the affinity-purified bait proteins (26). When Rix1-TAP, which is immobilized on CaM beads, was incubated with a MgCl 2 step gradient, Rea1 and 60 S subunit proteins were released with 100 mM salt (Fig. 1B, lane 1). However, Rix1, Ipi1, and Ipi3 remained bound under these conditions and were also resistant to 200 mM MgCl 2 , but they eluted by SDS-sample buffer (Fig. 1B, lane 3). Ipi1 reproducibly appeared substoichiometric after salt washing. It remains to be shown whether Ipi1 is present in lower amounts in the Rix1 complex or is not stained effectively by Coomassie Blue. Similar results were obtained for salt-treated Ipi3-TAP purification (data not shown). We conclude that Rix1, Ipi1, and Ipi3 form a salt-stable complex, to which Rea1 and the nascent 60 S subunit are less tightly attached. Recently, Krogan et al. (27), in a large scale effort to isolate RNA processing complexes by TAP purification following a ultracentrifugation step to remove ribosomes, have reported a similar complex. Finally, since Rea1 was observed to be specifically enriched in the Rix1-TAP purification, we examined whether a synthetic lethal relationship exists between the rea1 ts mutants and the rix1-1 mutant (10). However, we did not observe any sl phenotype (data not shown).
Rea1 and the Rix1 Complex Members Exhibit a Nucleoplasmic Location-Since Rea1 and the Rix1 complex are co-enriched, we wanted to know whether their subcellular distribution is also similar. Previously, we showed that Rix1 is localized in the nucleus (5). Ipi1 and Ipi3 were genomically tagged with the GFP epitope at their C terminus, which ensures that protein expression remains under the control of the native promoter. Expression of the genomically N-terminally tagged GFP-Rea1 is under the GAL promoter due to difficulties in obtaining a fully functional C-terminal GFP fusion protein. 2 After confirming that the GFP tagging had no effect on the growth rate of the strains (Supplemental Fig. 1), we examined the yeast cells under the fluorescence microscope. Similar to Rix1, Rea1, Ipi1, and Ipi3 were localized throughout the nucleoplasm (Fig. 2). We also expressed the DsRed-tagged nucleolar protein Nop1 in the GFP-tagged strains, and we did not observe nucleolar concentration for any of the proteins as this can be judged by the lack of colocalization between the GFP and the DsRed signals (Supplemental Fig. 2). These results are in agreement with the previously reported localization of Ipi1 and Rea1 proteins (28) and Rix1 (5). Thus, Rea1 and the Rix1 complex could function in late nucleoplasmic maturation of pre-60 S subunits and/or their export to the cytoplasm. Taken together, the identified Rix1 complex is present together with the AAA-type ATPase Rea1 in a late pre-60 S particle that is located in the nucleoplasm.
Rea1 and the Rix1 Complex Are Required for Nuclear Export of the 60 S Subunit-We wanted to determine whether Rea1 and the components of the Rix1 complex are involved in late nucleoplasmic steps during 60 S subunit biogenesis. Previously, we showed that the rix1-1 ts mutant is strongly impaired in 60 S subunit export, but rRNA processing was not significantly affected (10). We followed different strategies to obtain conditional-lethal mutants of the Rea1 protein and Rix1 complex members. For REA1, both a repressible GAL1::REA1 construct and ts rea1 alleles were generated (see "Experimental Procedures"). For the essential Rix1, Ipi1, and Ipi3, conditional-lethal degron (td) mutants were used. These degron mutants target the proteins for rapid degradation in vivo upon shift to 37°C (15). Furthermore, we have confirmed that rea1-7 and rea1-21 ts mutants could be rescued by the presence of a plasmid carrying the REA1 wild-type gene (Supplemental Fig. 3A). In the case of the td mutants, we used the available RIX1 wild-type plasmid (10) to verify that the expression of the wild-type gene is sufficient to rescue the lethal phenotype of rix1-td mutant (Supplemental Fig. 3B).
The Rpl25-GFP reporter assay was developed to monitor 60 S export in vivo (18,29). In wild-type cells, Rpl25-GFP is incorporated into nascent pre-60 S ribosomes and is rapidly exported to the cytoplasm. In all ts mutants generated (rea1-21, rix1, ipi1, and ipi3), the reporter protein was cytoplasmic at steady state (Fig. 3). Similar results were obtained with the rea1-7 mutant (data not shown). At early time points upon shift to the restrictive condition, the td mutants exhibited a mixed phenotype with cells accumulating the Rpl25-GFP protein either at their nucleolus or throughout their nucleoplasm (Fig.  3A, 1-h shift). At a later time point, all td mutants exhibited strong nuclear accumulation, whereas the nucleolus was either similar to nucleoplasmic staining or else devoid of any signal (Fig. 3A, 2-h shift; Supplemental Fig. 4). A similar assay for nuclear export of the small subunit showed cytoplasmic localization of Rps2-GFP comparable with wild-type in all of the mutants (Fig. 3). Both the rea1-21 ts and the repressible 2 K. Galani, unpublished data.

FIG. 2.
Nuclear location of GFP-tagged Rix1, Ipi1, Ipi3, and Rea1 in yeast cells. Rix1, Ipi1, and Ipi3 were tagged with GFP at the C terminus, maintaining their native promoter; Rea1 was tagged at the N terminus, and its expression is under the control of GAL1 promoter. The in vivo location of the indicated tagged proteins was analyzed in the fluorescence microscope. For microscopic inspection, cells were grown to mid-log phase, mounted on a microscopic slide, and photographed. GAL::REA1 mutants accumulated the Rpl25-GFP in their nucleoplasm upon shift to restrictive conditions (Fig. 3, B and C). Interestingly, in the case of the rea1-21 ts mutant, ϳ20% of the cells accumulated the Rpl25-GFP protein at their nuclear periphery after 9 h of shift. The same phenotype was also observed with the rea1-7 ts mutant, albeit to a lesser extent (15%; data not shown). The fact that pre-60 S ribosomes accumulated in the entire nucleoplasm or at the nuclear periphery (Fig. 3, rea1-21 insert; Supplemental Fig. 5) suggests that a late step in nuclear export from the nucleoplasm to the cytoplasm is blocked in rea1 and rix1 complex mutants.
Rea1 and the Rix1 Complex Are Required for Normal Levels of the 60 S Subunit-As the export is impaired for Rea1 and the Rix1 complex mutants, we wished to see whether the overall production of 60 S subunits is likewise reduced. For this reason, we analyzed the ribosomal and polysomal profiles in rea1 and rix1 complex mutants by sucrose density gradient centrifugation. This analysis revealed a significant reduction of 60 S subunits as compared with 40 S in the mutants. Moreover, the appearance of half-mer polysomes was observed in mutant strains (Fig. 4), indicating a lack of mature 60 S to bind to the 43 S preinitiation complex. These data are consistent with the export defect, demonstrating that it is specific to the 60 S pathway.
Depletion of Rea1 or the Rix1 Complex Inhibits Synthesis of the 5.8 S rRNA-To determine whether the rea1 and rix1 complex member mutants are impaired in pre-rRNA process-ing, we performed Northern hybridization and primer extension analyses. The locations of oligonucleotide probes are indicated in Fig. 6A. Analyses of low molecular weight RNAs (Fig.  5A, upper panel) revealed that genetic depletion of Rea1 resulted in strong inhibition of processing from 7 S pre-rRNA to 5.8 S ϩ 30 and 6 S pre-rRNA, as shown by their substantial loss 16 h after transfer of the GAL::REA1 strain to glucose medium. Since the 7 S pre-rRNA was mildly accumulated, whereas the normal products of its processing were drastically reduced, it is likely that much of the 7 S pre-rRNA is degraded, presumably reflecting degradation of the entire pre-60 S particle. It is, however, difficult to directly assess this experimentally since 7 S is only faintly visible in pulse-chase labeling of wild-type strains, and the 5.8 S ϩ 30 and 6 S pre-rRNAs are not observed (data not shown). Mature 5.8 S was reduced relative to the tRNA 3 Leu loading control, as was the 5 S rRNA component of the 60 S subunit. Analyses of high molecular weight RNA by Northern hybridization (Fig. 5A, middle panel) and primer extension (Fig. 5A, lower panel) showed the accumulation of the 35 S pre-rRNA and 23 S RNA, accompanied by a mild reduction in the 20 S and 27 SA 2 pre-rRNA. Very similar processing defects were observed in the rea1-21 ts strain 4 h after transfer to the non-permissive temperature (data not shown). These phenotypes indicate a delay in early pre-rRNA processing at sites A 0 , A 1 , and A 2 (Fig. 6B), which are very frequently seen in strains defective in 60 S subunit synthesis (30). The 27 SB pre-rRNAs were mildly accumulated, with little alteration in FIG. 3. Nuclear export of 60 S subunits is inhibited in mutants of the Rix1 complex and Rea1. A and B, the indicated wild-type cells (wt-td, REA1) and degron (rix1-td, ipi1-td, ipi3-td) or rea1-21 ts mutants were either grown at 23°C (permissive temperature) or shifted to 37°C (restrictive temperature) for 1 and 2 h (wt-td, rix1-td, ipi1-td, ipi3-td) or 9 h (rea1-21). C, the GAL1::REA1 repression strain was grown in galactose-containing medium (permissive condition) or shifted for 14 h to glucose-containing medium (restrictive condition). As further indicated in the figure, the strains expressed either the 60 S subunit reporter Rpl25-GFP or the 40 S subunit reporter Rps2-GFP. Subcellular location of the Rpl25-GFP and Rps2-GFP reporters was analyzed by fluorescence microscopy. Note that Rpl25-GFP tends to accumulate at the nuclear periphery in rea1-21 cells (in particular in the daughter nucleus of dividing cells) after 9 h at 37°C (inset picture), whereas Rps2-GFP remains cytoplasmic. the levels of processing at the alternative B 1S and B 1L sites, indicating that the early steps in 60 S synthesis continue in the Rea1-depleted strain.
Related defects in 5.8 S synthesis were seen in strains depleted of Rix1, Ipi1, or Ipi3 (Fig. 5B, upper panel). In each case, the 7 S pre-rRNA was accumulated relative to the wild-type control following transfer to non-permissive conditions, accompanied by reduced 6 S pre-rRNA. The level of the 5.8 S ϩ 30 was reduced in the wild-type following transfer to 37°C but was further reduced in the mutant strains. Some reduction in the levels of the mature 5.8 S was seen after 4 h in nonpermissive conditions. The mature rRNAs are not generally turned over and are therefore depleted only by growth under non-permissive conditions. Greater depletion would not therefore have been expected over this period. Few clear alterations were seen in the processing of the earlier, high molecular pre-rRNAs. However, some increase in the level of 35 S was seen in the rix1-td strain at later time points, whereas 23 S accumulated in the ipi1-td strain, both indicative of a mild delay in the early pre-rRNA processing steps (Fig. 5B,  middle panel).
We conclude that depletion of Rea1, Ipi1, Ipi3, or Rix1 each inhibits the processing of the 7 S pre-rRNA to 5.8 S ϩ 30 and 6 S rRNA accompanied by degradation of the pre-60 S particle, with consequent reduced synthesis of 5.8 S rRNA. Mild defects seen in early pre-rRNA processing steps are probably indirect, as many other mutations leading to impaired 60 S subunit synthesis have been reported to have similar effects (30). DISCUSSION Biogenesis of ribosomal subunits is a complicated process, which requires the spatial and temporal coordinated function of greater than 170 trans-acting factors. A major challenge now is to assign functions to these many components, determine their nearest neighbor relationships and organization in subcomplexes, group them in classes with similar functions, and map their positions within the structure of the nascent pre-60 S subunit.
In this study, we focused on a late nucleoplasmic pre-60 S intermediate that is close to export to the cytoplasm. This particle contains the 560-kDa AAA-type ATPase Rea1, which is distantly related to the motor protein dynein, and the Rix1 complex, which is composed of Rix1, Ipi1, and Ipi3 subunits. Mutants of these non-ribosomal factors result in the defective 3Ј processing of the 5.8 S rRNA, a late pre-rRNA processing step, and export of the 60 S subunit from the nucleoplasm to the cytoplasm.
Rea1 is predicted to be the largest yeast protein at 560 kDa and belongs to the family of AAA ϩ -type ATPases (31). Members of this family are involved in diverse processes such as membrane fusion, proteolysis, DNA replication and recombination, microtubule organization, and intracellular motility (32,33). All the functions of the AAA family can be linked to its central property of affecting protein-protein interactions. Their common feature is the presence of an AAA-ATPase module, which is required for ATP binding and hydrolysis. Structural analysis FIG. 4. Production of 60 S subunits is blocked in mutants of the Rix1 complex and Rea1. The indicated wildtype cells (wt-td) and degron (rix1-td, ipi1-td, ipi3-td) or rea1-21 ts mutants were either grown at 23°C or shifted to 37°C (restrictive condition) for 6 h (wt-td, rix1-td, ipi3-td, rea1-21) or 4 h (ipi1-td). The GAL::REA1 conditional mutant was grown in galactose-containing medium before shifting to glucose-medium for 16 h. Analysis of ribosome and polysome profiles (A 254 nm ) was then performed by sedimentation centrifugation on sucrose density gradients. Arrows indicate 40 S, 60 S, and 80 S ribosomes, polysomes, and "half-mer" polysomes.
revealed that the AAA domain forms hexameric or heptameric rings after oligomerization, which change their conformation depending on the bound nucleotide (34,35). Instead of the more common one or two AAA protomers per molecule, Rea1 possesses six domains, similar to the AAA motor protein dynein. Detailed sequence analysis of the AAA protomers, in both dynein and Rea1, demonstrate a closer relatedness to each other than to any other AAA protein, suggesting that they evolved from a common ancestor (13). If Rea1 performs a function related to dynein, it could be involved in intranuclear movement or export of pre-60 S particles from the nucleoplasm to the cytoplasm.
In addition to the N-terminal AAA motifs, Rea1 has a MIDAS domain located at the carboxyl terminus. The MIDAS FIG. 5. Depletion of Rea1 and the Rix1 complex members inhibits pre-rRNA processing. A, the strains were pregrown in galactose medium and then transferred to glucose medium for the times indicated. Upper panel, Northern hybridization analyses of low molecular weight RNA separated on a 8% polyacrylamide gel containing urea. Middle panel, Northern hybridization analyses of high molecular weight RNA separated on a 1.2% agarose gel containing urea. Lower panel, primer extension analyses. Probes used are indicated on the left of each panel. wt, wild type. B, the strains were pregrown in raffinose medium containing CuSO 4 at 23°C (R lanes). Galactose was added to induce Ubr1 expression at 23°C (23 lanes), and strains were transferred to medium lacking CuSO 4 at 37°C for 1-4 h (lanes 1-4). Panels and probes are as described for A, except for the addition of a further loading control, scR1 RNA, which is the RNA component of the cytoplasmic signal recognition particle.
domain, most notably present in integrins, is involved in protein-protein interactions. Interestingly, deletion of the C terminus of Rea1 containing the MIDAS motif is lethal 2 .
Rea1 represents the second AAA ATPase shown to have a role in ribosome biogenesis. Rix7 was the first such protein found to be required for large subunit formation (11). RNA processing analyses indicate that the rix7-1 mutation (11) inhibits an earlier step in the biogenesis process than do the rea1-21 mutant or Rea1 depletion. We examined whether the rix7-1 mutation was synthetic lethal with either of the rea1 ts mutant alleles (data not shown), but no sl phenotype was observed.
Rea1 and the Rix1 complex are required for 3Ј maturation of the 5.8 S rRNA, which is inhibited in rea1 and rix1 complex mutants. No accumulation of intermediates between 7 S and 6 S pre-rRNA were observed in strains depleted of Rea1, Ipi3, Ipi1, or Rix1. This is in contrast to the effects of depletion of components of the exosome complex, which is believed to be directly responsible for this processing reaction. This suggests that in the absence of Rea1 and the Rix1 complex members, 3Ј processing cannot be initiated from site C 2 , perhaps because the 3Ј end of the pre-rRNA is sequestered in the pre-rRNA structure (Fig. 6B). It is notable that in the predicted structure of ITS2, the C 2 cleavage site is indeed predicted to lie within a base-paired region (Ref. 36 and references therein).
They reported overall defects in RNA processing with specific reduction in 25 S rRNA, 20 S pre-rRNA, and 18 S rRNA and accumulation of 7 S levels. It is possible that these more general effects are due to the extended 24-h depletion time course used. Here we used degron constructs that result in rapid protein depletion and observed defects in the ITS2 processing within 1 h of shifting to restrictive conditions. Our data suggest that depletion of the Rix1 complex proteins alters the structure of a late nuclear pre-60 S particle, leading to inhibition of the 3Ј processing of 7 S pre-rRNA and degradation of the pre-ribosome.
We also report that Rea1 and the Rix1 complex members are required for the export of the large ribosomal subunit. This defect is apparently specific as small subunit export was unimpaired. Strikingly, the ts mutants of rea1 accumulated Rpl25-GFP at the nuclear periphery, suggesting an involvement just prior to or during export. This would be consistent with Rea1 being responsible for removal of proteins that retain the nascent ribosome in the nucleus.
A possible explanation for our results is that the ATPase activity of the Rea1 complex is used for the structural remodeling of the pre-ribosome, including the separation of the 5Ј and 3Ј regions of ITS2. The processing of ITS2, triggered by the Rea1/Rix1 complex-induced pre-60 S remodeling, could act as a quality control step required to allow maturation of 60 S subunits into export-competent particles. Delays in ITS2 processing have also been reported in strains defective for other late- acting factors required for pre-60 S export, including Nmd3, Gsp1, and Rrp12 (37)(38)(39).
Reduction of the levels of two of the putative GTPases, Nog1 and Nug2, which were found associated with the Rea1/ Rix1 complex containing pre-60 S particles also exhibits defects in ITS2 processing. Disruption of Nog1 function through RNA interference led to a dramatic decrease in the levels of free 60 S particles and the appearance of an atypical rRNA intermediate in which ITS2 was not cleaved (40). Likewise, depletion of Nug2 resulted in a dramatic decrease in 5.8 S rRNA levels and accumulation of the 7 S precursor as well as nucleolar accumulation of ITS2-containing precursors (6). These GTPases may act in concert with the Rea1 ATPase to coordinate the structural reorganization of late pre-60 S ribosomes, prior to the irreversible step of nuclear export.