A Conserved SET Domain Methyltransferase, Set11, Modifies Ribosomal Protein Rpl12 in Fission Yeast*

SET domain-containing methyltransferases post-translationally modify a variety of cellular proteins, such as histones, cytochrome c, ribulose-bisphosphate carboxylase/oxygenase, and ribosomal proteins. In the fission yeast Schizosaccharomyces pombe, at least 13 SET domain-containing proteins have been identified in the genome, four of which are involved in transcriptional regulation through their modification of histone tails. However, the roles played by the other SET domain proteins in cellular processes and their physiological substrates remain unresolved. We show here that S. pombe Set11, a SET domain-containing protein encoded by SPCC1223.04c, specifically modifies Rpl12 (ribosomal protein L12). Recombinant Set11 prepared from Escherichia coli had catalytic activity and methylated a 17-kDa polypeptide in cellular extracts of set11 mutant cells. The methylated protein was isolated by two-dimensional gel electrophoresis or by reverse-phase chromatography and was identified as Rpl12 by mass spectrometry. In vitro methylation experiments using wild-type and mutant Rpl12 proteins verified that Set11 modified recombinant Rpl12 and suggested that its potential target site was lysine 3. The methylation site modified by Set11 was also confirmed by mass spectrometric analysis, which also revealed other unique methylation sites of Rpl12. Finally, we found that Set11 predominantly localized to the nucleolus and that the overproduction of Set11 caused a severe growth defect. These results suggest that Rpl12 methylation occurs during the ribosomal assembly processes and that control of the Set11 expression level is important for its cellular function.

SET domain-containing methyltransferases post-translationally modify a variety of cellular proteins, such as histones, cytochrome c, ribulose-bisphosphate carboxylase/oxygenase, and ribosomal proteins. In the fission yeast Schizosaccharomyces pombe, at least 13 SET domain-containing proteins have been identified in the genome, four of which are involved in transcriptional regulation through their modification of histone tails. However, the roles played by the other SET domain proteins in cellular processes and their physiological substrates remain unresolved. We show here that S. pombe Set11, a SET domaincontaining protein encoded by SPCC1223.04c, specifically modifies Rpl12 (ribosomal protein L12). Recombinant Set11 prepared from Escherichia coli had catalytic activity and methylated a 17-kDa polypeptide in cellular extracts of set11 mutant cells. The methylated protein was isolated by two-dimensional gel electrophoresis or by reverse-phase chromatography and was identified as Rpl12 by mass spectrometry. In vitro methylation experiments using wild-type and mutant Rpl12 proteins verified that Set11 modified recombinant Rpl12 and suggested that its potential target site was lysine 3. The methylation site modified by Set11 was also confirmed by mass spectrometric analysis, which also revealed other unique methylation sites of Rpl12. Finally, we found that Set11 predominantly localized to the nucleolus and that the overproduction of Set11 caused a severe growth defect. These results suggest that Rpl12 methylation occurs during the ribosomal assembly processes and that control of the Set11 expression level is important for its cellular function.
The post-translational modification of proteins is a ubiquitous mechanism by which cells expand the function of proteins. The SET domain was originally identified in three Drosophila proteins, Su(var)3-9, Enhancer of zest, and Trithorax, which are all involved in epigenetic gene regulation (1) and were later demonstrated to be histone lysine methyltransferases (2,3). Although specific lysine residues in histone tails are generally acknowledged as substrates for the SET domain-containing methyltransferases (4), this domain has also been found in a subset of methyltransferases that modify non-histone proteins, such as ribulose-bisphosphate carboxylase/oxygenase (5), cytochrome c (6), p53 (7,8), and ribosomal proteins (9).
The methylation of ribosomal protein has been identified in a diverse range of species, including bacteria, yeast, and humans. Escherichia coli L11, one of the most well characterized ribosomal proteins, is ␣-N-trimethylated at Ala 1 and ⑀-N-trimethylated at Lys 3 and Lys 39 (10). These methyl groups are added by a single methyltransferase called PrmA (11). The methylation positions and PrmA are highly conserved among bacterial clades, suggesting an important function, although PrmA is dispensable for normal growth (12,13). In the budding yeast Saccharomyces cerevisiae, direct mass spectrometric analysis of the large ribosomal proteins revealed that six of them, Rpl1, Rpl3, Rpl12, Rpl23, Rpl42, and Rpl43, are post-translationally modified by the addition of methyl groups (14). Rpl23 is specifically ⑀-N-dimethylated at two residues, Lys 105 and Lys 109 , and these modifications are catalyzed by the SET domain-containing methyltransferase Rkm1 (15,16). S. cerevisiae Rpl12, the counterpart of bacterial L11, is also modified,: by ⑀-N-dimethylated at Lys 3 , ⑀-N-trimethylation at Lys 10 , and ␦-N-monomethylation at Arg 66 (17,18). Rkm2, another SET domain-containing protein, and Rmt2, a protein arginine methyltransferase, catalyze the methylation at Lys 10 and Arg 66 of Rpl12, respectively (17,18). The enzyme(s) responsible for the ⑀-N-dimethylation of Lys 3 in Rpl12 has yet to be identified. In mammals, several mass spectrometric studies have identified methyl modifications on ribosomal proteins (19 -22). In most cases, however, the precise methylation sites and responsible enzyme(s) have yet to be explored. Currently, only one mammalian ribosomal methyltransferase (Prmt3) has been shown to modify a ribosomal protein, S2 (23). Although ribosomal protein methylation appears to be conserved among different organisms, the physiological role(s) of the methyl modification is poorly understood.
In the fission yeast Schizosaccharomyces pombe, at least 13 SET domain-containing proteins have been identified in the genome. Our previous studies showed that the fission yeast Clr4, a homolog of Su(var)3-9, is a histone H3 Lys 9 -specific methyltransferase that is required for heterochromatin assembly (24). Although several other SET domain-containing proteins in fission yeast add methyl groups to specific lysine residues of histone tails (25)(26)(27), the roles played by other SET domain proteins in cellular processes and their physiological substrates remain unresolved. Furthermore, although S. pombe Rmt3, a protein-arginine methyltransferase, has been shown to modify the small ribosomal protein S2 (Rps2) and to be involved in stability of the small subunit (28,29), little attention has been paid to the involvement of SET domain-containing proteins in ribosomal biogenesis.
In this study, we analyzed S. pombe Set11 and found that it is responsible for the methylation of ribosomal protein L12 (Rpl12). Interestingly, although S. pombe Set11 shows a higher sequence similarity to S. cerevisiae Rkm2, which is responsible for modifying Rpl12 at Lys 10 (18), it preferentially catalyzed ⑀-N-trimethylation at Lys 3 of Rpl12. We further confirmed the methylation site modified by Set11 by mass spectrometric analysis, which led us to identify other unique methylation sites of Rpl12. Finally, we showed that Set11 predominantly localized to the nucleolus in S. pombe, suggesting that it functions specifically in ribosome assembly.

EXPERIMENTAL PROCEDURES
Strains and Plasmids-The strains used in this study are listed in supplemental Table S1. All of the yeast strains were grown at 30°C in YEA (0.5% yeast extract, 3% glucose, 75 g/ml adenine) or minimal medium EMM (MPP Biomedicals) supplemented with amino acids (75 g/ml adenine, leucine, and histidine and 150 g/ml uracil). The deletion and tagging of endogenous genes were conducted using a PCR-based genetargeting protocol (30) and YEA medium containing antibiotics (200 g/ml G418 or hygromycin). To construct plasmids for producing recombinant Set11 or Rpl12 proteins in E. coli, the coding sequence for set11 ϩ or rpl12 ϩ was amplified by PCR and cloned into the pRSET (Invitrogen) and pTriEX-4 Hygro (Novagen) vectors, respectively. To obtain plasmids for expressing mutant recombinant Rpl12 protein in E. coli, the above plasmids were subjected to site-directed mutagenesis as described (31). To express Set11 and EGFP-fused 3 Set11 in S. pombe, the set11 ϩ coding sequence was cloned into the multi-ple cloning site of pREP1 (a multicopy episomal plasmid carrying the nmt1 promoter) or a pREP1 derivative containing the EGFP-coding sequence (pREP1-EGFP), and the resulting plasmids were designated pREP1-set11 ϩ and pREP1-EGFP-set11 ϩ , respectively. To obtain S. pombe strains expressing Set11 with short internal deletions (⌬NHSP189 -192 and ⌬GEQIFLCY216 -224), the set11 ϩ -coding sequence was first cloned into the pCR2.1-TOPO plasmid (Invitrogen), and each deletion was introduced by in vitro site-directed mutagenesis (31). After insertion of a ura4 ϩ marker gene, the resultant plasmid was digested with NruI and introduced into the original set11 ϩ locus of the SPYB109 strain by isolating ura4 ϩ -expressing cells. To replace the wild-type SET domain with the mutated domain, strains that lost the ura4 ϩ gene by internal homologous recombination were isolated using counter-selective medium containing 5-fluoroorotic acid. The deletions of the set11 ϩ coding region were confirmed by PCR. All other strains were constructed using standard genetic crosses.
Preparation of HeLa and Yeast Cell Extracts-To prepare histone-enriched HeLa nuclear extracts, HeLa cells grown to ϳ80 -90% confluence in 100-mm dishes were washed twice with phosphate-buffered saline and then lysed in 1 ml (per dish) of nuclear lysis buffer (100 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1.5 mM MgCl 2 , 0.65% Nonidet P-40, 10 mM butyrate, pH 8.0, 1 mM phenylmethylsulfonyl fluoride) supplemented with proteinase inhibitor mixture (Complete TM EDTA-free; Roche Applied Science). The nuclei were harvested by scraping into a 1.5-ml microcentrifuge tube and collected by centrifugation at 500 ϫ g for 5 min at 4°C. After removal of the supernatant, the nuclei were resuspended in 200 l of 0.4 N H 2 SO 4 , and the suspension was further incubated with rotation for 2 h at 4°C. The insoluble fraction was removed by centrifugation at 5,500 ϫ g for 5 min at 4°C. The acid-soluble supernatant was collected into a new centrifuge tube, and the proteins were precipitated with 50 l of 100% trichloroacetic acid (final concentration, 20%). The suspension was placed on ice for 1 h and spun at 22,000 ϫ g for 15 min at 4°C. The protein pellet was washed once with acidified acetone (0.1-0.3% HCl) and twice with acetone, air-dried, and resuspended in 100 l of deionized H 2 O by vortexing. After the recovery was estimated, the extracted proteins were stored at Ϫ80°C until use. S. pombe nuclear extracts of the wild-type or ⌬set11 mutant cells were prepared as described previously (32).
Two-dimensional Electrophoretic Analysis of Proteins-The two-dimensional gel analysis of methylated proteins was performed as described previously (33).
Chromatographic Fractionation of Proteins-S. pombe cell extracts prepared as described above were loaded onto a SOURCE 15RPC ST 4.6/100 column (GE Healthcare) that had been equilibrated with Eluent A (0.065% trifluoroacetic acid in water) using a liquid chromatography system (AKTAexplore 10S; GE Healthcare). The bound proteins were eluted with a 2-100% linear gradient of Eluent B (0.055% trifluoroacetic acid in acetonitrile). The proteins in each fraction were dried under a vacuum, dissolved in deionized H 2 O, and subjected to the in vitro methyltransferase assay.

Analysis of Methylated Peptide by Nano-liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS)-
The proteins in each gel slice were subjected to reduction with 10 mM dithiothreitol, alkylation with 55 mM iodoacetamide, and digestion with 10 g/ml modified trypsin (Promega), 20 g/ml ArgC (Sigma), or 20 g/ml GluC (Sigma) at 37°C for 16 h. After the in-gel digestion, the peptides were extracted with 5% formic acid and 50% acetonitrile, dried under a vacuum, and dissolved in 2% acetonitrile and 0.1% formic acid. The multiple digested peptides were then fractionated by C18 reverse-phase chromatography (Paradigm MS4; Microm BioResources) and applied directly into a quadrapole ion trap mass spectrometer (Finni- gan LTQ; Thermo Fisher Scientific) with a Fortis tip mounted on a three-dimensional stage (AMR, Tokyo, Japan). The ion trap was programmed to carry out three successive scans consisting of, first, a full-scan MS over the range 450ϳ2000 m/z, and second and third, data-dependent scans of the top two abundant ion obtained in the first scan. Automatic MS/MS spectra were obtained from the highest peak in each scan by setting a relative collision energy of 35% and exclusion time of 15 min for molecules of the same m/z value range. The molecular masses of the resulting peptides were searched against the nonredundant NCBI data base using the MASCOT program.
Fractionation of Ribosomes-Fractionation of ribosomes from the wild-type and ⌬set11 mutant S. pombe strains was performed as described previously (34).
Microscopy Analysis-To analyze the localization of EGFP-fused Set11, wild-type and ⌬set11 mutant S. pombe cells were transformed with the pREP1-EGFP-set11 ϩ plasmid. The transformed cells were cultured on plates with minimal medium lacking leucine (AA-leu). Single colonies were picked and patched onto new plates. The cells were grown to early log phase in liquid medium and washed twice with deionized H 2 O, and the DNA was visualized by incubation with 1 g/ml Hoechst 33342. Microscopic images were captured on a Zeiss Axioplan 2 imaging microscope and an ORCA-ER camera (Hamamatsu).

RESULTS
S. pombe set11 ϩ Encodes an Active Methyltransferase-In the fission yeast S. pombe, at least 13 genes that potentially encode SET domain-containing proteins (SET proteins) have been identified in the genome (Table 1). Among these genes, four (set1 ϩ , set2 ϩ , clr4 ϩ , and set9 ϩ ) encode histone lysine methyltransferases (24 -27). Although several of the encoded SET proteins have homologous counterparts in other organisms (35) (Table 1), the function of the remaining nine SET proteins is unclear. To elucidate the roles of the SET proteins in S. pombe, we started by analyzing the enzymatic activity of these uncharacterized proteins using an in vitro methyltransferase assay. Based on a previous genome-wide protein localization analysis (36), we initially examined Set3, Set11, and Set13, because they are localized predominantly to the nucleus (Table 1). We prepared recombinant His-tagged  proteins for full-length Set11 (His-Set11) and Set13 (His-Set13), and the N-terminal half of Set3 (His-Set3N) ( Fig. 1A and data not shown), and incubated each of these recombinant proteins with histone-enriched HeLa nuclear extract in the presence of a 3 H-labeled methyl donor ([ 3 H]AdoMet). Recombinant His-tagged Clr4 (His-Clr4) was also tested as the control for histone H3-K9 methyltransferase (24). Although no activity was detected for His-Set13 and His-Set3N under these conditions, we found that His-Set11 possessed an intrinsic methyltransferase activity and that it specifically methylated a protein(s) with a molecular mass of ϳ17 kDa on SDS-PAGE (Fig.  1B, p17me). Because no methylated band was detected in the mock experiments (Fig. 1B, mock), it is unlikely that the methylation of p17 was catalyzed by endogenous methyltransferases. These results led us to focus on the physiological roles and substrate(s) of the Set11 methyltransferase.
Because the methylated protein(s) appeared to migrate slightly more slowly than that of histone H3 methylated by His-Clr4 (Fig. 1B), it was likely that Set11 target protein(s) was a minor histone species or other nuclear protein. To characterize this substrate of Set11, we performed the same in vitro methyltransferase assay using histone-enriched nuclear extracts prepared from wild-type or ⌬set11 mutant S. pombe strains. Although specific methylation signals were not detected in the assay using wild-type cellular extract, a strongly methylated band with a similar molecular mass as in the HeLa nuclear extract (ϳ17 kDa) was observed in the assay using the ⌬set11 cellular extract (Fig. 1C, p17me). These results suggested that p17 is a physiological target of Set11 methyltransferase, and that its methylation site(s) may have been modified already by endogenous Set11 in the wild-type cells. In addition, these results imply that p17 and its target site(s) are evolutionarily conserved from fission yeast to humans.
Set11 Modifies the Ribosomal Protein Rpl12 in S. pombe-To identify the target protein(s) of the Set11 methyltransferase, we first fractionated the methylated product(s) by two-dimensional acetic acid-urea-Triton X-100 (AUT) and acetic acid-urea-cetyltrimethylammonium bromide (AUC) gel analysis, which is generally used to analyze histone species (33). After separation on the first (AUT) and second (AUC) gels, two discrete signals were detected in the autoradiograph ( Fig. 2A). These two protein species migrated to almost the same level in the first AUT gel but showed a different migration pattern in the second AUC gel. The difference might be attributable to the presence or absence of modification(s) other than the methylation by Set11. The protein spots corresponding to these signals were excised from the two-dimensional gel and subjected to LC-MS/MS analysis. In parallel with this two-dimensional gel analysis, we also used chromatographic approaches to separate the substrate(s) of Set11 protein. Nuclear extracts prepared from ⌬set11 mutant strains were fractionated by reverse-phase chromatography, and the eluted proteins were tested in the in vitro methyltransferase assay. As shown in Fig. 2B, the target protein(s) was eluted in several fractions with a peak at fraction 22. The protein band showing the same elution profile in the chromatography was excised and subjected to LC-MS/MS analysis. From both the two-dimensional gel and chromatographic approaches, we obtained a series of peptides that matched perfectly with the deduced amino acid sequence of S. pombe ribosomal protein L12 (Rpl12) (Fig.  2C). Rpl12, a counterpart of bacterial L11 protein, is a highly conserved protein from yeast to humans (Fig. 2C). It was recently shown that Rpl12 in S. cerevisiae is methylated at Lys 10 by a SET domain-containing protein, Rkm2 (18).
Recombinant Rpl12 Is Methylated in Vitro by Set11-To confirm that Rpl12 is a physiological substrate for Set11, an in vitro methyltransferase assay was performed using recombinant full-length and C-terminal-deleted Rpl12 proteins (Fig. 3A, Rpl12-His and Rpl12⌬C-His). We found that fulllength Rpl12-His was clearly methylated by Set11 and that Rpl12⌬C-His was a better substrate for the Set11 methyltransferase activity (Fig. 3B). These results indicated that Rpl12 is a physiological substrate for Set11 and demon- strated that the methylated residue(s) resides in the N-terminal 1-106 residues of Rpl12. The role played by the C-terminal region of Rpl12 in the methyltransferase reaction is currently unclear, although it is possible that steric interactions between the N-and C-terminal domains of Rpl12 affect the ability of Set11 to methylate its target site. Because we failed to produce an N-terminally deleted recombinant Rpl12 mutant in E. coli, we cannot exclude the possibility that a target residue(s) of Set11 is also present in the C-terminal region of Rpl12. However, previous studies on E. coli L11 (10) and S. cerevisiae L12 (18) and our mass spectrometric analysis support the idea that a specific lysine residue in the N-terminal region of Rpl12 is the preferred target site of Set11 (see below).
To characterize further the site of Rpl12 methylation, we introduced a series of Ala substitutions for the candidate Lys residues in Rpl12⌬C-His (Fig. 3A, indicated by red) and used these mutant proteins in the in vitro methyltransferase assay (Fig.  3C). We found that several combinations of the Ala substitution greatly affected the Set11 methyltransferase activity (see Rpl12⌬C-His K3, 10    was clearly methylated by Set11 (Fig. 3C). Taken together, these results suggest that Rpl12 is a physiological substrate of Set11 and that the Lys 3 , and potentially Lys 39 and Lys 40 , of Rpl12 are the candidate target residues for Set11 activity.
Determination of the Methylation Sites of Rpl12 by MS/MS-To determine the in vivo methylation sites of Rpl12, the endogenous Rpl12 in wild-type (set11 ϩ ) or ⌬set11 mutant cells was isolated by reverse-phase chromatography (as shown in Fig.  2B). The corresponding band was excised from an SDS-PAGE gel and digested with ArgC or GluC peptidase, which cleaves preferentially on the C-terminal side of arginine or glutamate residues. The digested peptide fragments were then analyzed using LC-MS/MS. The overall elution profiles of ArgC-digested Rpl12 peptides in the nano-LC spectra were superimposed, and the masses of representative peaks were matched between the wild-type and ⌬set11 mutant cells (Fig. 4, A and D). However, a relatively abundant peptide was eluted in a prominent peak at 13.20 min in the ⌬set11 mutant (Fig. 4D, asterisk) that was not observed in the wild-type cells. MS/MS analysis revealed that the amino acid sequence of this fragment matched the N-terminal residues 1-15 of Rpl12 with the initial Met removed (Fig. 4F). The experimental mass of this peptide calculated from the mean m/z of MH 3ϩ was 1848.59 Ϯ 0.32 (MH ϩ ) ( Table 2). The corresponding 1-15 peptide for Rpl12 in wild-type cells was eluted at 13.34 min (Fig. 4, A, asterisk, and C), and its experimental mass was 1890.37 Ϯ 0.49 (MH ϩ ) ( Table 2). Of particular interest was that the mass difference between these wild-type and ⌬set11 mutant peptides was ϳ41.78 Da, which corresponds to the mass of three methyl groups. Although there are two potentially methylated lysine residues in this peptide at Lys 3 and Lys 10 , the MS/MS results (Fig. 4, C and F) and parallel experiments using GluC, in which a similar ϳ42-Da difference was observed for the N-terminal 1-8 peptide (supplemental Fig. S1 and Table 2), strongly suggest that that the Rpl12 in wild-type cells is trimethylated at Lys 3 , and this methylation is absent in the ⌬set11 mutant cells. This is quite consistent with our observation obtained from the in vitro methyltransferase assay (Fig. 3C).
Of note, the mass of the N-terminal 1-8 peptide even in the ⌬set11 mutant was still 28 -29 Da larger than that of the theoretical mass of the corresponding peptide (Table 2; experimental mass, 972.04 (MH ϩ ) and theoretical mass, 943.45). Although the margin of mass error needs to be considered, the additional mass appears to correspond to that of two methyl groups. In addition, MS/MS results using ArgC or GluC suggested that the additional mass could be localized to the first three amino acids, aside from the ⑀-N-trimethylation at Lys 3 . These observations support the idea that the N-terminal Proline 1 (Pro1) is modified by ␣-N-dimethylation after removal of the initial Met, as previously reported for S. cerevisiae Rpl25 (37). By analyzing the MS/MS results, we were also able to identify a monomethylation at Arg 66 (supplemental Fig. S1C) and a methylation and/or another modification(s) at Lys 39 and Lys 40 ( Table 2). The most frequently observed mass for the latter region was a dimethylation at Lys 39 (supplemental Table S2). We found that these modifications were the same in the ⌬set11 mutant cells. We could not obtain clear evidence for the methyl modification at Lys 10 , which is observed in S. cerevisiae Rpl12 and is the exclusive target of Rkm2 (18).
Functional Analysis of Set11 in S. pombe-To investigate the enzymatic function of Set11 in vivo, we introduced short internal deletions (⌬NHSP189 -192; ⌬NHSP and ⌬GEQIFLCY 216 -224; ⌬GE-Y) into the two conserved regions of the core SET domain in Set11 (2). Nuclear extracts prepared from wildtype or mutant strains (⌬set11, ⌬set11NHSP, and ⌬set11GE-Y) were then subjected to the in vitro methyltransferase assay (Fig.  5A). Rpl12 from the ⌬set11NHSP and ⌬set11GE-Y strains was clearly methylated by recombinant His-Set11 to the same extent as that from ⌬set11, indicating that these deletions abolished the in vivo enzymatic activity of the endogenous Set11  proteins. These results demonstrate that the two conserved regions of the SET domain core are important for the catalytic site for Set11 methyltransferase activity.
To obtain further insight into the physiological role of Set11, we examined its localization by expressing EGFP-fused Set11 protein in wild-type or ⌬set11 mutant cells. Although the previous study described a nuclear localization of Set11 (36), EGFP-Set11 predominantly localized to the nucleolus, which coincides with the nuclear hemisphere that stained weakly with Hoechst 33342, in both wildtype and ⌬set11 mutant cells (Fig.  5B). These results suggest that the Lys 3 methylation of Rpl12 occurs during the ribosome assembly process in the nucleolus and support the idea that, if Set11 has substrates other than Rpl12, they are probably also involved in ribosomal assembly or function.
S. pombe cells in which the set11 gene was completely deleted or that expressed catalytically inactivated Set11 protein were viable and showed no growth defects under a variety of conditions (data not shown). Bacterial ribosomal protein L11, the counterpart of eukaryotic Rpl12, binds a highly conserved domain of 23 S rRNA and is thought to be involved in the ribosomal GTPase activity of the dynamic decoding process (38). Although little is known about the roles played by L11 methylation in this process, the methyl modifications of Rpl12 could play a role in its association with the ribosome or in ribosomal function. To test these possibilities, we first examined whether the Rpl12 methylation affects its association with the ribosome. Mature ribosomes were isolated from wild-type and ⌬set11 mutant cells by sucrose density centrifugation, and the level of Rpl12 associated with the ribosomes was assayed by immunoblotting. We found, however, that the level of Rpl12 was not affected by the ⌬set11 mutation (Fig. 5C).
Next, to examine the potential role of Rpl12 methylation in ribo-somal function, we analyzed the sensitivity to cycloheximide of ⌬set11 mutant cells. Cycloheximide is a widely used compound that inhibits protein synthesis by blocking transla- tional elongation via interference with the peptidyl transferase activity of the 60 S ribosome. In S. pombe, two paralogous genes (rpl1201 and rpl1202) encode Rpl12, and mutant cells in which one of these genes was deleted (⌬rpl1202) showed higher cycloheximide sensitivity, suggesting that the dosage of Rpl12 is critical to ribosomal function (Fig. 5D). However, ⌬set11 cells showed no increased sensitivity to cycloheximide, suggesting that the Rpl12 methylation has little, if any, effect on the ribosomal function. In previous studies, it was shown that the loss of S. pombe Rmt3, the arginine methyltransferase for ribosomal protein S2 disrupts the balance of small 40 S and large 60 S subunits because of a deficit in the small subunit level (28,29). In line with this, we also examined the stability of the 40 S and 60 S subunits by polysome analysis. We observed, however, no clear difference between the wild-type and ⌬set11 mutant strains (data not shown), suggesting that Rpl12 methylation at Lys 3 is not directly involved in ribosomal subunit stability or that it has overlapping functions with other methyl modification(s). The set11 ϩ gene is located immediately downstream of another gene, spc19 ϩ (exactly 20 bp away from the spc19 ϩ open reading frame), and our Northern blotting and reverse transcription-PCR analyses showed that set11 ϩ mRNA is transcribed polycistronically with spc19 ϩ mRNA (data not shown). Previous studies have indicated that the expression of set11 ϩ is regulated by cell cycle and meiotic processes (39,40). To examine whether the expression level of set11 ϩ is important for its function, we overproduced Set11 in wild-type cells (Fig.  5E). Interestingly, we found that the induction of set11 ϩ expression by the nmt1 promoter caused a severe growth defect (Fig. 5F, Set11-OP). Because the level of Rpl12 in Set11-overproducing cells was comparable with that in wild-type cells (Fig. 5G), it is unlikely that the growth defect was simply caused by a change in Rpl12 stability. Although it remains to be tested whether this effect is attributable to the disturbance in Rpl12 modification, these results indicate that set11 ϩ expression is tightly regulated, and its level is critical for proper cellular function.

DISCUSSION
In this study, we used an in vitro methyltransferase assay and identified Rpl12 as a physiological substrate of Set11, which is the first SET domain-containing methyltransferase in S. pombe discovered to modify a non-histone protein. Although the targets of S. cerevisiae Rkm1 and Rkm2 were identified by a combination of in vivo labeling of mutant strains for potential SET methyltransferases and mass spectrometric analysis (15,18), our in vitro methyltransferase assay appears to be useful for identifying physiological substrate(s) of uncharacterized SET domain methyltransferases. Indeed, by applying this approach, FIGURE 5. In vivo enzymatic activity and subcellular localization of Set11. A, the SET domain core is essential for the Set11 enzymatic activity. S. pombe nuclear extracts prepared from wild-type and set11 mutant strains (⌬set11, ⌬set11NHSP, and ⌬set11GE-Y) were resolved by 15% SDS-PAGE and visualized by Coomassie staining (left panel). These extracts were incubated with His-Set11 and [ 3 H]AdoMet. The proteins were resolved by 15% SDS-PAGE, and labeled proteins were detected by autoradiography (right panel). The positions of size markers and the methylated Rpl12 are indicated. B, EGFP-tagged Set11 predominantly localizes to the nucleolus. EGFP-tagged Set11 was expressed from the nmt1 promoter in wild-type (bottom row) or ⌬set11 cells (top row). The cells were stained with Hoechst33342 (DNA). Merged images of EGFP-Set11 and DNA (Merged), differential interference contrast (DIC), and schematic drawing of the cells are also shown. C, the level of Rpl12-EGFP associated with the ribosome is not affected by set11 ϩ deletion. Total cell extracts (W) of wild-type and ⌬set11 cells were fractionated by discontinuous sucrose density centrifugation into a post-ribosomal supernatant (S), a sucrose cushion (C), and pellet (P). The fractions were analyzed by immunoblotting using antibodies against green fluorescent protein and Rpl19 (control) proteins. Total RNA was extracted, resolved by 0.8% agarose gel, and stained with ethidium bromide. Most of the ribosomes were fractionated in the sucrose cushion (C) and pellet (P) fractions as indicated by an asterisk. D, deletion of set11 ϩ does not affect cycloheximide sensitivity. 5-fold dilutions of wild type, ⌬set11, ⌬rpl1202, and ⌬rpl1202⌬set11 strains were plated onto YEA alone and YEA containing different doses of cycloheximide. E-G, overproduction of Set11 causes a growth defect but does not change the level of Rpl12. Wild-type cells were transformed with pREP1 or pREP1-set11 ϩ , and 5-fold dilutions of the transformed cell cultures were plated onto minimal medium lacking leucine (AA-Leu) (F). The level of Set11 (E) and Rpl12-3HA (G) was analyzed by immunoblotting using antibodies against Set11 and the hemagglutinin tag, respectively. we have identified the substrate of Set13 methyltransferase as another ribosomal protein. 4 Considering that S. pombe Set8 and Set10 show similarity with Rkm1 (Table 1), it is highly likely that they also target and modify ribosomal proteins. If this is the case, at least four S. pombe SET methyltransferases, Set8, Set10, Set11, and Set13, are involved in the modification of ribosomal proteins. In this respect, one hypothesis to explain this is that a subfamily of SET methyltransferases may have evolved to target the lysine residues of highly basic proteins, such as histones and ribosomal proteins, that stably associate with DNAs and RNAs, respectively (16). Although our results demonstrate that Rpl12 is a physiological substrate of Set11 methyltransferase, it is also possible that Set11 has other substrate proteins. We have searched for potential target proteins by using different fractions of cellular lysate. However, we have yet to identify any other target proteins for Set11 (data not shown). Considering the nucleolar localization of Set11, it is likely that its other substrate proteins would also be involved in ribosomal biogenesis. Our results indicate that Rpl12 is methylated at five residues: ␣-N-dimethylation at Pro 1 , ⑀-N-trimethylation at Lys 3 , ⑀-N-dimethylation at Lys 39 and/or Lys 40 , and ␦-N-monomethylation at Arg 66 and thus underscore the evolutionary conservation of Rpl12 methylation (Fig. 6). Although we showed that Set11 modifies Lys 3 , and it is likely that an Rmt2 homolog (SPAC26A3.17c) catalyzes Arg 66 methylation, the responsible enzymes for the other methyl modifications remain unclear. Detailed mass spectrometric analyses of Rpl12 treated with candidate SET methyltransferases and their mutants will help identify the responsible enzyme(s) for the Lys 39 /Lys 40 methylation.

E. coli
N-terminal methylation has been described for only a small number of proteins (41). Of ribosomal proteins, S. cerevisiae Rps25 was recently shown to be dimethylated at the ␣-N-terminal proline after cleavage of the initial methionine (37). Interestingly, in the N-terminal sequence of S. cerevisiae Rps25, [M]PPKQQ-, the first three residues are the same as those of Rpl12 ([M]PPKFD-). In addition, human RCC1 has also been shown to have ␣-N-terminal methylation (42). Although the methylating enzyme has yet to be determined, mutational analysis of human RCC1 revealed that [M]-(S/P/A)-P-K serves as a substrate recognition motif for N-terminal methylation, which is consistent with yeast Rpl12 and Rps25, implying that there are conserved mechanisms for N-terminal methylation. In most cases, the physiological function of N-terminal methylation is unclear. However, intriguingly, methylation-defective mutants of RCC1 bind less effectively to chromatin during mitosis, which causes a spindle pole defect (42). Therefore, it is possible that the N-terminal methylation of Rpl12 regulates its interaction with ribosomal RNA.
The physiological function of the Set11 methyltransferase is unclear. Yeast cells lacking the enzyme are viable, and no particular defect in ribosomal function was observed in ⌬set11 cells. The same is true for the ⌬rkm2 mutant budding yeast cells (18) and ⌬prmA mutant E. coli cells (12). Because Rpl12 possesses multiple methyl modifications, it is possible that these modifications act cooperatively, and the roles played by the methyl modifications may become clear when set11 is combined with other mutations in the relevant SET domain methyltransferases. We have shown that Set11 predominantly localizes to the nucleolus, suggesting that the methylation of Rpl12 at Lys 3 occurs during the ribosome assembly processes. The dynamics and stability of this methyl modification are not known. Because Rpl12 prepared from wild-type cells was not a good substrate for recombinant Set11 in vitro, it is likely that Rpl12 is predominantly methylated at Lys 3 in wild-type cells. Interestingly, we found that the overproduction of Set11 caused a severe growth defect (Fig. 5E). In Caenorhabditis elegans, Rpl12 regulates its own splicing, and the overproduction of Rpl12 increases the proportion of unproductively spliced mRNAs (43). A similar alternative splicing mechanism has been suggested for the production of mammalian Rpl12 (44). Although the molecular mechanisms underlying the effect of Set11 overproduction are currently unclear, it is possible that Rpl12 or another Set11 target protein(s) is involved in controlling the levels of ribosomal proteins through its RNA binding properties, and methyl modification may modulate this regulatory function. It is also possible that the autoregulation system of ribosomal biogenesis as observed for C. elegans Rpl12 may suppress the effect of the ⌬set11 mutation. These hypotheses will be tested in future studies that take ribosome homeostasis into consideration.