Rrp17p Is a Eukaryotic Exonuclease Required for 5′ End Processing of Pre-60S Ribosomal RNA

Summary Ribosomal processing requires a series of endo- and exonucleolytic steps for the production of mature ribosomes, of which most have been described. To ensure ribosome synthesis, 3′ end formation of rRNA uses multiple nucleases acting in parallel; however, a similar parallel mechanism had not been described for 5′ end maturation. Here, we identify Rrp17p as a previously unidentified 5′–3′ exonuclease essential for ribosome biogenesis, functioning with Rat1p in a parallel processing pathway analogous to that of 3′ end formation. Rrp17p is required for efficient exonuclease digestion of the mature 5′ ends of 5.8SS and 25S rRNAs, contains a catalytic domain close to its N terminus, and is highly conserved among higher eukaryotes, being a member of a family of exonucleases. We show that Rrp17p binds late pre-60S ribosomes, accompanying them from the nucleolus to the nuclear periphery, and provide evidence for physical and functional links between late 60S subunit processing and export.

sections, treated as previously described (Oeffinger et al., 2007). A MALDI orthogonal time of flight (prOTOF 2000, Perkin Elmer Science) and a MALDI linear ion trap mass spectrometer (vMALDI LTQ, Thermo Fisher Scientific) were used for peptide fingerprinting (MS) and amino acid sequencing (MS/MS), respectively (Krutchinsky, 2001). XProteo (www.xproteo.com) was used to correlate peptide mass fingerprint data or tandem MS CID data obtained from MS and MS/MS analyses and enabled identification of proteins (Ossipova et al., 2006).

Cosedimentation and Velocity Gradient Analysis
Sucrose gradient centrifugation was performed as described (Baßler et al., 2001;Tollervey et al., 1993). RNA was extracted from each fraction and resolved on standard 1.2% agarose/formaldehyde gel. Mature rRNAs and pre-rRNA species were detected by ethidium staining and Northern hybridization, respectively. Sedimentation of proteins was assayed by SDS-PAGE and PrA-tagged Rrp17p was detected by Western immunoblotting with peroxidase-conjugated rabbit IgG (SIGMA).
Velocity Centrifugations were carried out on a 5-20% (w/w) sucrose gradient as described in (Alber et al., 2007). The markers used were aprotinin, cytochrome C, carbonic anhydrase and bovine serum albumin (Sigma). Briefly, 5µg of recombinantly purified Rrp17p and marker proteins were centrifuged on a sucrose/TB (20mM Hepes, pH7.5, 110mM KOAc, 0.1% Tween 20, 1mM DTT, 1:200 Solution P) gradient at ~300,000 g max for 24h in an SW55 Ti rotor at 4 °C. 200μl fractions were collected from the top of the gradient and analyzed by SDS-PAGE and Coomassie staining. Band intensities were quantified using ImageJ, and the peak fractions of the marker proteins were plotted as function of its sedimentation coefficient and fitted with a standard curve by linear regression (r-squared >0.99 in all cases).

RNA Mobility Shift and Exonuclease Assays
Labeled and unlabeled RNA substrates were synthesized in vitro by T7 polymerase transcription of pBluescript(+) linearized with XbaI, and of rDNA (5'-ITS1 to 3'-ITS2) that had been amplified from an rDNA plasmid and is carrying the T7 promoter region. RNAs were purified on 15% acrylamide/urea gels, excised from the gel, eluted and precipitated. The RNA mobility shift binding reaction was performed in 30mM Tris-HCl pH 7.4, 150 mM KCl, 2mM MgCl 2 , 0.1% Triton X-100, 20% glycerol and 1mM DTT, in the presence of tRNA (1mg/ml), 0.25pmol of 32 P-labelled pre-rRNA and 0-200nmol of recombinant protein in a reaction volume of 15 µl. RNA was heat denatured at 65°C for 10 min, followed by slow cooling to room temperature, and then added to the binding reaction. Reactions were incubated at room temperature for 30 min and then loaded on a 6% native acrylamide/bisacrylamide (80:1)/4% glycerol gel in 0.5x TBE buffer. Prior to loading, the gel was pre-run for 1 h and then run for 3h at 250 V in the cold room (Fatica et al., 2002b).
Exonuclease assays were performed in 10mM Tris-HCl pH 7.6, 50mM KCl, 1mM MgCl 2 , 10mM DTT, 100µg/ml BSA and 0.8U/µl in the presence of 0.5pmol 32 P-labelled mRNA and 50mM of recombinant protein in 15µl total reaction volume. Samples were incubated at room temperature for 0-15min and then loaded onto 20% acrylamide/urea gels in 1x TBE buffer and run for 2hs at 200V.

Fluorescent Microscopy, In Situ Hybridization, and Cell Imaging
For fluorescent in situ hybridization (FISH), cells growing in permissive or shifted to non-permissive medium for 12hr were fixed and hybridized with pre-rRNA probes as described in (Zenklusen et al., 2008). ITS1 probe (TGGACTCTCCATCTCTTGACT TCTTGCCCAGTAAAAGCTCTCATGCTCTT) labeled with Cy5 and ITS2-1 probe (ATAGGCCAGCAATTTCAAGTTAACTCCAAAGAGTATCACTC) labeled with Cy3 were used.
Bold letters indicate labeled nucleotides.
All images were acquired using a Olympus BX61 wide-field epi-fluorescence microscope using an Olympus 100x, 1.35NA objective with HC DIC optics from cells grown in synthetic medium lacking histidine. Multiple fields of cells were counted (~300) and % of cells determined displaying phenotype.

Poison Assay
GFP-tagged strains were grown to early-mid-log phase (A600 = 0.4) for steady-state analysis and then treated essentially as described by (Shulga et al., 1996) using metabolic energy poisons to arrest ATPdependent steps. 10 ml of cells were harvested at room temperature, washed with 1 ml ddH20; 500µl of washed cell suspension were pelleted again and resuspended in 1 ml of 10 mM sodium azide ("azide"), 10 mM 2-deoxy-o-glucose ("deoxyglucose") in glucosefree SC medium and incubated for 30 min at 30°C. Poison-treated cells and untreated cells were then imaged using a Olympus BX61 wide-field epifluorescence microscope using an Olympus 100x, 1.35NA objective.

Figure S6. Expression of Recombinant Rrp17p Wild-Type and Mutant Proteins
Rrp17p WT (A) and mutant proteins (B, C) were expressed in E.coli as (His) 10 fusions and eluted from an NTA column using imidazol. (D) Constructs carrying wildtype RRP17 (pRS414-3xHA-RRP17-TRP), NOL12 (hRrp17) or different point and truncation mutants were transformed into strains carrying a RRP17 deletion. Efficient expression of plasmid-expressed wild-type, mutant proteins and the loading control PGK1 (E) were tested by Western Blot analysis using anti-HA and anti-PGK1 antibodies on cell lysates from strains grown at 37°C to mid-log phase.

Processing
Localization of pre-60S (ITS2-1) and pre-40S (ITS1) ribosomal subunits. (A) An xrn1Δ strain was grown to mid-log phase. Cells were fixed and mounted with DAPI to stain the nuclei (blue). An increase of 5'ITS1 signal in the cytoplasm was observed in xrn1Δ cells, due to the accumulation of excised ITS1 fragments in the absence of Xrn1p. A normal nucleolar distribution of ITS2-1 was observed. (B) GAL::rrp17/Met::rat1 cells exhibited normal nucleolar distribution of ITS2-1 signal n permissive medium. After 12 hr in nonpermissive medium, ITS2-1 was distributed throughout the entire nucleus. In the presence of Xrn1p, no increased ITS1 signal was observed in the cytoplasm, but a normal nucleolar distribution was detected. Bars represent 10µm. (C) Localization of ITS2-1 and ITS1 in rai1Δ and rrp6Δ cells. The strains were grown to mid-log phase. Cells were fixed and mounted with DAPI to stain the nuclei (blue). In the absence of either Rai1p (left) and Rrp6p (right), cells ITS2-1 was distributed throughout the entire nucleus, while ITS1 exhibited a normal nucleolar distribution. Bars represent 10µm.