Proteomic Analysis of Human Nop56p-associated Pre-ribosomal Ribonucleoprotein Complexes

Nop56p is a component of the box C/D small nucleolar ribonucleoprotein complexes that direct 2′-O-methylation of pre-rRNA during its maturation. Genetic analyses in yeast have shown that Nop56p plays important roles in the early steps of pre-rRNA processing. However, its precise function remains elusive, especially in higher eukaryotes. Here we describe the proteomic characterization of human Nop56p (hNop56p)-associated pre-ribosomal ribonucleoprotein complexes. Mass spectrometric analysis of purified pre-ribosomal ribonucleoprotein complexes identified 61 ribosomal proteins, 16 trans-acting factors probably involved in ribosome biogenesis, and 29 proteins whose function in ribosome biogenesis is unknown. Identification of pre-rRNA species within hNop56p-associated pre-ribosomal ribonucleoprotein complexes, coupled with the known functions of yeast orthologs of the probable trans-acting factors identified in human, demonstrated that hNop56p functions in the early to middle stages of 60 S subunit synthesis in human cells. Interestingly, the nucleolar phosphoprotein treacle, which is responsible for the craniofacial disorder associated with Treacher Collins syndrome, was found to be a constituent of hNop56p-associated pre-rRNP complexes. The association of hNop56p and treacle within the complexes was independent of rRNA integrity, indicating a direct interaction. In addition, the protein compositions of the treacle-associated and hNop56p-associated pre-ribosomal ribonucleoprotein complexes were very similar, suggesting functional similarities between these two complexes with respect to ribosome biogenesis in human cells.

The ribosome constitutes one of the most fundamental molecular machines in living cells. Given that protein synthesis is essential for cell growth, proliferation, and adaptation, ribo-some biogenesis is intimately coupled to the needs of the cell. Ribosome biogenesis is efficiently coordinated in the nucleolus, a subnuclear compartment in eukaryotic cells. Each mature ribosome in the cytoplasm consists of the large subunit (60 S) and the small subunit (40 S) that together comprise over 80 ribosomal proteins organized within and around mature rRNAs. In mammalian cells, the large subunit is composed of ϳ50 ribosomal proteins and three species of rRNA (28 S, 5.8 S, and 5 S), whereas the small subunit consists of ϳ30 ribosomal proteins and 18 S rRNA. During ribosome biogenesis, ribosomal DNA for 5.8 S, 18 S, and 28 S rRNAs is transcribed by RNA polymerase I into a large primary precursor (47 S) containing 5Ј-and 3Ј-external transcribed spacers (5Ј-ETS 1 and 3Ј-ETS) and two internal transcribed spacers (ITS1 and ITS2) (1,2). Concomitant with the methylation and pseudouridylation of ribose moieties, the 47 S precursor is cleaved at specific sites to produce a series of characteristic intermediates that ultimately result in mature 5.8 S, 18 S, and 28 S rRNAs. Pre-5 S rRNA is transcribed by RNA polymerase III and processed independently of the other three rRNAs (3).
During ribosome biogenesis, a multitude of small nucleolar RNAs (snoRNAs) and trans-acting proteins, neither of which are contained in mature ribosomes, form pre-ribosomal RNP (pre-rRNP) complexes and play crucial roles in the processing and modification of pre-rRNAs as well as the assembly of rRNAs with ribosomal proteins (4 -7). Genetic analyses in yeast over the past 3 decades have identified a number of the trans-acting proteins that are involved in ribosome biogenesis, and their ordered association with and dissociation from pre-rRNP complexes have been outlined (5)(6)(7). However, the precise functions of these trans-acting factors remain obscure. Furthermore, our knowledge of the function of these factors in mammalian cells has been greatly limited by difficulty in identifying mammalian rRNP components because of limited sample availability and the inability to conduct routine genetic analyses.
Proteomic methodology has been facilitated by advances in tag-based purification methods for protein complexes as well as the development of highly sensitive mass spectrometric techniques and search engines that access huge amounts of genomic data. The advent of proteomic analyses allows the characterization of large complexes consisting of Ͼ100 proteins (8,9). In fact, proteomic analysis of yeast pre-rRNP complexes has proven highly fruitful for the characterization of transacting factors involved in ribosome biogenesis (10 -13). We recently purified and characterized the nucleolin-and the parvulin-associated pre-rRNP complexes, both of which are involved in human ribosome biogenesis (14,15). These studies clearly demonstrated the applicability of proteomic analysis to the study of human ribosome biogenesis, and also identified a number of mammalian counterparts of yeast trans-acting factors as well as additional trans-acting factors that had not been previously identified in yeast.
Here we present a proteomic analysis of Nop56p-associated pre-rRNP complexes in human cells. Nop56p is a member of the core proteins of box C/D snoRNP complexes (16,17) that direct 2Ј-O-methylation of pre-rRNA ribose moieties during the early stages of pre-rRNA processing (16, 18 -23). The box C/D snoRNAs are characterized by their conserved box C (RUG-AUGA) and box D (CUGA) motifs that are essential for function (18, 24 -27). In addition to Nop56p, the core proteins Nop58p, Nop1p (fibrillarin), and Snu13p (a 15-kDa protein in human) form the mature box C/D snoRNP complex. The box C/D snoRNP complex has been reconstituted in vitro, and fibrillarin is responsible for its methylation activity (21). Mouse Nop56p is 36% identical and 58% similar to Nop58p at the amino acid level (17). Because only a single ortholog of Nop56p/Nop58p has been identified in Archaea, these two proteins most likely diverged from the archaeal protein (17). Genetic analysis of yeast Nop56p indicates that it is responsible for cleaving 35 S pre-rRNA (the primary precursor in yeast) at sites A 0 , A 1 , and A 2 to produce 25 S rRNA (the mature rRNA corresponding to human 28 S rRNA) (16). The C-terminal region of yeast Nop56p contains repeated KK(E/D) motifs (16) that are also present in the yeast pseudouridine synthetase Cbf5p (22), and a similar motif (KKX) is found in the yeast putative ATP-dependent RNA helicase Dbp3p (27). However, the KK(E/D) motifs are not essential for Nop56p function in yeast, and therefore the relevance of this motif to Nop56p activity is not clear. Human Nop56p (hNop56p) lacks these specific motifs but contains a stretch of highly basic residues within its C-terminal region (16,17). Although Nop56p is known to play a significant role as a protein trans-acting factor in ribosome biogenesis in yeast, the precise function of this protein remains unknown, especially in mammalian cells (16). Our present data demonstrate that, within hNop56p-associated pre-rRNPs, hNop56p interacts directly with treacle, the Treacher Collins Syndrome (TCS) gene product.
Epitope-tagged hNop56p and Treacle Vectors and Expression in 293EBNA Cells-A DNA fragment encoding hNop56p with the FLAG tag (DYKDDDDK) at its N terminus was amplified by PCR using a human placenta cDNA library template (OriGene Technologies, Inc., Rockville, MD) with the primers 5Ј-ATATAGCTAGCGCCACCATGGA-CTACAAGGACGACGACGACAAGGCGGGACGTGGCGCCATGGT-G-3Ј and 5Ј-TATATGGATCCTTATTTTCTTCCTTCGACCAT-3Ј. For treacle, a DNA fragment encoding human treacle with the FLAG tag at its C terminus was amplified as above using the primer set 5Ј-TATA-GCTAGCGCCACCATGGCCGAGGCCAGGAAGCGGCGG-3Ј and 5Ј-T-ATAGGATCCTCACTTGTCGTCGTCGTCCTTGTAGTCTACAGTCTG-CTCTGCTGTCTTC-3Ј. The hNop56p and treacle fragments were cut with the NheI and BamHI, respectively, and ligated into the corresponding restriction sites of the expression vector pcDNA3.1(ϩ) (Invitrogen).
Human 293EBNA cells were grown in DMEM supplemented with 10% heat-inactivated fetal calf serum, streptomycin (0.1 g/ml), and penicillin G (100 units/ml) at 37°C in an incubator with an atmosphere of 5% CO 2 . Subconfluent cells in 90-mm dishes were transfected individually with 10 g of the appropriate expression plasmid using Lipo-fectAMINE according to the manufacturer's instructions, and the transfected cells were grown for 48 h at 37°C.
Isolation of hNop56p-and Treacle-associated Complexes-At 48 h post-transfection, 293EBNA cells were harvested and washed with PBS and then lysed in lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% IGEPAL CA630, 1 mM phenylmethanesulfonyl fluoride) on ice for 30 min. The soluble fraction was obtained by centrifugation at 15,000 rpm for 30 min at 4°C and was incubated with 20 l of anti-FLAG M2-agarose with gentle mixing overnight at 4°C to immunoprecipitate hNop56p-or treacle-associated complexes. After washing the agarose five times with lysis buffer and then once with 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, the bound complexes were eluted with 20 l of 50 mM Tris-HCl, pH 7.5, 150 mM NaCl containing 500 g/ml of the FLAG peptide. The isolated complexes were analyzed by SDS-PAGE on 12.5% gels.
Ribonuclease Treatment of hNop56p-associated Complexes-Immunoprecipitated hNop56p-associated complexes (bound to anti-FLAG M2-agarose) were incubated in 50 mM Tris-HCl, pH 7.5, 150 mM NaCl containing 10 g/ml RNase A for 10 min at 37°C, washed twice with lysis buffer and once with 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and eluted with the FLAG peptide in the same manner described above.
Immunocytochemistry-293EBNA cells were grown on 8-well culture slides (Biocoat, BD Biosciences). Subconfluent cells were transfected with the expression plasmids using LipofectAMINE as above. Prior to fixation, the cells were washed with PBS, followed by incubation with 3.7% formaldehyde in PBS. After several washes with PBS-T (PBS containing 0.05% (w/v) Tween 20), the cells were incubated with PBS containing 0.1% (w/v) Triton X-100 for 5 min at room temperature. The cells were then blocked using 3% (w/v) non-fat dried milk in PBS and were incubated with anti-FLAG for 1 h at room temperature. The cells were washed with PBS-T and then incubated with Alexa Fluor 488conjugated anti-mouse IgG for 1 h at room temperature, followed by three washes with PBS-T. The cells were then examined with a Fluo-View confocal laser-scanning microscope (Olympus, Tokyo, Japan).
Protein Identification by Peptide Mass Fingerprinting-SDS-PAGE gel fragments containing proteins were cut out and subjected to in-gel protease digestion as described (29). Peptides generated by the digestion were recovered and analyzed for peptide mass fingerprint using a PE Biosystems MALDI-TOF/MS (Voyager DE-STR) as described (29). Peptide masses were searched with 50 ppm mass accuracy using the data base fitting program MS-Fit (prospector.ucsf.edu), and protein identification was carried out using the criteria described previously (15).

Protein Identification by LC-MS/MS Analysis-hNop56p-associated
complexes were digested with proteases, and the resulting peptides were analyzed by an LC-MS/MS system at the femtomole level as described (30). The peptide mixture was separated on a frit-less Mightysil-C18 (3-m particle, Kanto Chemical, Osaka Japan) column (45 ϫ 0.150 mm inner diameter) using a gradient of acetonitrile (0 -40%) in 0.1% formic acid over 80 min at a flow rate of either 50 or 25 nl/min by direct nano-LC system (Nanosolution, Tokyo, Japan). The eluted peptides were sprayed directly into a quadrupole time-of-flight hybrid mass spectrometer (Q-Tof2, Micromass, Manchester, UK). MS/MS spectra were acquired by data-dependent collision-induced dissociation, and the MS/MS data were analyzed using MASCOT software (Matrix Science, Wyndham Place, UK) for peptide assignment.

RESULTS
Purification of hNop56p-associated RNPs-To purify hNop56p-associated complexes, we constructed an expression plasmid encoding hNop56p fused to the FLAG epitope (FLAG-hNop56p). In 293EBNA cells, transiently overexpressed FLAG-hNop56p localized primarily to the nucleolus (Fig. 1A). hNop56p-associated complexes that formed in FLAG-hNop56ptransfected cells were purified via immunoprecipitation using anti-FLAG-conjugated beads. SDS-PAGE indicated that the hNop56p-associated complexes contained more than 50 proteins spanning a broad molecular weight range (Fig. 2, lane 2). Of these, four proteins were also immunoprecipitated from non-transfected control cells (Fig. 2, lane 4) thus reflecting nonspecific binding. RNase treatment of the complexes while bound to the immunoprecipitating beads resulted in dissociation of almost all protein constituents except for FLAG-hNop56p and its degradation products (Fig. 2, lane 3). These results indicated that the purified complexes were ribonucleoproteins (RNPs) and that the association of the protein components was RNA-dependent.
Protein Components of hNop56p-associated RNPs-The protein components of the immuno-isolated hNop56p-associated RNP complexes were identified by mass spectrometry. Protein identification was first carried out for individual bands excised from SDS-PAGE gels using in-gel digestion and MALDI-TOF/ MS. This analysis yielded 35 ribosomal proteins (29 large subunit proteins and 6 small subunit proteins) and 17 non-ribosomal proteins (Fig. 3,and Supplemental Material Tables I-III).
We described previously a highly sensitive direct nano-flow LC-MS/MS system to identify proteins in limited quantities of multiprotein complexes (30). Therefore, intact immuno-isolated hNop56p-associated RNPs were digested with A. lyticus protease I, and the digest was analyzed directly by nano-LC-MS/MS. More than 2,500 MS/MS spectra were obtained, from which ϳ500 peptides were assigned to 104 proteins using the Mascot search software. Of these, 61 were ribosomal proteins (42 large subunit proteins and 19 small subunit proteins) and 43 were non-ribosomal proteins (Tables I and Table II, and  Supplemental Material Tables I-III). This LC-MS/MS analysis identified all the proteins documented by the electrophoresisbased analyses except three. A total of 107 proteins were identified in the hNop56p-associated RNP complex using the electrophoresis-and LC-based analyses. Of these, 62 were ribosomal proteins and 45 were non-ribosomal proteins (Tables  I and II, and Supplemental Material Tables I-III).
A functional comparison of the 45 non-ribosomal proteins with yeast orthologs revealed that 16 proteins represent putative trans-acting factors involved in ribosome biogenesis (Table  I). These proteins were classified into three functional groups, namely box C/D snoRNP proteins, RNA helicases, and other proteins. Together with Nop56p, fibrillarin, Nop5/Nop58p, and non-histone chromosome protein 2-like 1 (equivalent to a 15.5-kDa protein) form the core of box C/D snoRNPs that direct 2Ј-O-methylation of pre-rRNAs at specific sites. One of the box C/D snoRNPs, U3, processes yeast pre-rRNAs at the A 0 , A 1 , and A 2 sites (16, 18 -20). Two largest class of trans-acting factors involved in ribosome biogenesis in yeast (7). The hypothetical protein FLJ10377, peter pan homolog, EBNA1-binding protein 2, BRIX, DKFZP564M182 protein, and a putative nucleotide-binding protein (estradiol-induced) have not been reported to function in ribosome biogenesis in mammals although their yeast counterparts are constituents of pre-ribosomal particles ( Table I). The roles of nucleolin and B23 in pre-rRNA processing are partly established in mammals. Nucleolin plays multiple roles in early processing of rRNAs and in the packaging of nascent rRNAs (31)(32)(33)(34). B23 is involved in the late stage of rRNA processing (35). Nucleolar and coiled-body phosphoprotein 1, designated Nopp130 in human and Nopp140 in mouse, associates with snoRNPs required for rRNA modification and processing and appears to function as a chaperone for ribosome biogenesis and intranuclear transport (36,37). Heterogeneous nuclear ribonucleoprotein U (hnRNP U) is implicated in RNA metabolism because it interacts with other protein components of hnRNP complexes (38) as well as the C-terminal domain of RNA polymerase II (39). Its yeast ortholog, Nop3p, is an essential nucleolar protein, and its depletion leads to pre-rRNA processing defects at the middle stage of 60 S ribosome subunit biogenesis (40). The identification of these putative trans-acting factors strongly suggested that the hNop56p-associated RNP is involved in ribosome biogenesis.
In addition to the 16 trans-acting factors, 29 non-ribosomal proteins that are not known to function in ribosome biogenesis were identified in hNop56p-associated RNPs. Among these, at least 21 proteins exhibited nucleolar and/or nuclear localization (Table II), suggesting that they may indeed be involved in ribosome biogenesis.
Pre-rRNA Species in hNop56p-associated RNPs-Given that our proteomic analysis indicated that hNop56p-associated RNPs might participate in ribosome biogenesis, the RNP complexes were examined for the appropriate rRNA species. For this purpose, RNAs were isolated from the complexes and subjected to Northern blot analysis using oligonucleotide probes corresponding to the 5Ј-ETS, 18 S and 5.8 S, ITS2, and 28 S regions of 47 S pre-rRNA (Fig. 4A). The pre-rRNA species 47 S, 45 S, 41 S, 32 S, 30 S, and 26 S were identified (Fig. 4B). Because the hNop56p-associated RNPs also contained 16 possible trans-acting factors involved in ribosome biogenesis, we concluded that this RNP constituted de facto pre-ribosomal particles (pre-rRNP complexes). In addition to the pre-rRNAs, the complexes also contained mature rRNAs (5.8 S, 18 S, and 28 S), whereas neither pre-rRNAs nor mature rRNAs were detected in the corresponding preparation isolated from untransfected 293EBNA cells (data not shown). This suggested that the mature rRNAs were endogenous components of the hNop56passociated complexes. We reported previously that human parvulin-associated pre-rRNP complexes also contain mature rRNAs even though the complexes were purified from the nuclear fraction (15). Thus, the presence of the mature rRNAs in these complexes suggests that pre-rRNA maturation may occur within the nucleolus or nucleus in vivo and/or possibly during the purification of the complexes in vitro.
hNop56p Interacts Directly with Treacle, the Treacher Collins Syndrome Gene Product-No clear protein bands were observed on the SDS gel of the RNase-treated hNop56p-associated pre-rRNPs, except for those of FLAG-hNop56p and proteolytic fragments thereof (Fig. 2, lane 3). However, when the same preparation was analyzed directly by LC-MS/MS without gel separation, one particular protein, treacle, was identified specifically (i.e. seven peptide hits; see Supplemental Material Fig. 1). The two independent LC-MS/MS analyses also identified this protein in hNop56p-associated pre-rRNPs (14 and 18 peptide hits, respectively, as shown in Fig. 5 and Supplemental Material Fig. 1 as well as Table II and Supplemental Material  Table III). Thus, treacle remained associated with hNop56p even after RNase treatment of the hNop56p-associated pre-rRNPs, suggesting that it binds directly to hNop56p in a manner that is independent of rRNA integrity. These data provided the first evidence of a physical and perhaps functional relationship between hNop56p and treacle, the Treacher Collins syndrome gene (TCOF1) product that is responsible for a craniofacial disorder (41)(42)(43). Treacle contains 1411 amino acids with a calculated molecular mass of 144,228 Da. However, this protein reportedly migrates as a diffuse smear at ϳ220 kDa during SDS-PAGE, due presumably to hyper-phosphorylation (44). Thus, the fact that treacle was not identified by our gel-based MALDI-TOF/MS analysis of the hNop56p-associated pre-rRNPs (Fig. 3) most likely reflects its anomalous migration during SDS-PAGE.
Treacle-associated RNP Complexes Contain hNop56p-Treacle is thought to belong to a class of nucleolar phosphoproteins on the basis of its amino acid sequence (42,43) and its subcellular localization within the nucleolus (45). To confirm that this protein is indeed a component of hNop56p-associated pre-rRNPs, a reverse pull-down experiment was performed in which treacle-associated complexes were isolated from 293EBNA cells using FLAG-tagged treacle as bait. Exogenously expressed FLAG-tagged treacle localized to the nucleolus (Fig. 1, C and D) and migrated at ϳ220 kDa during SDS-PAGE ( Fig. 6; identification by MALDI-TOF/MS analysis: 36 peptides matched; 27% peptide coverage; mean error ϭ 1.68 ppm). These data suggested that FLAG-tagged treacle, like the endogenous protein (44), became phosphorylated when expressed in 293EBNA cells. The protein profiles of the treacleassociated and hNop56p-associated complexes were very simi-  Fig. 2. Each protein band was excised, subjected to in-gel protease digestion, and analyzed by the MALDI-TOF/MS. The resulting peptide mass fingerprints were matched to a data base, and the identity of each protein is shown to the right of the gel. Lane 1, molecular weight markers; lane 2, proteins of the hNop56p-associated complexes. lar (Fig. 6). In addition, MALDI-TOF/MS analysis of treacleassociated complexes identified hNop56p (11 peptides matched; 21% peptide coverage; mean error ϭ Ϫ8.61 ppm) as well as a number of components found in common with the hNop56-associated pre-rRNP, including hnRNP U (10 peptides matched; 16% peptide coverage; mean error ϭ Ϫ6.73 ppm), nucleolin (21 peptides matched; 32% peptide coverage; mean error ϭ Ϫ8.31 ppm), and B23 (nine peptides matched; 30% peptide coverage; mean error ϭ Ϫ20.17 ppm) (Fig. 6). We also confirmed that the association of all these protein components with treacle-associated complexes required RNA integrity (using RNase as described for hNop56-associated pre-rRNPs; data not shown). Thus, it is likely that treacle, like hNop56p, is contained within an RNP complex and participates in certain stages of ribosome biogenesis in human cells. DISCUSSION We purified hNop56p-associated complexes from 293EBNA cells via transfection with a FLAG-tagged hNop56p construct. hNop56p-associated complexes were shown to be pre-rRNPs  based on the fact that the complexes contained pre-rRNAs as well as many ribosomal proteins and potential trans-acting factors involved in ribosome biogenesis. Because the pre-rRNP complexes associated with hNop56p contained the 47 S primary precursor rRNA that corresponds to the yeast 35 S rRNA, the hNop56p-associated RNP possibly contains a pre-ribosome particle formed at a very early stage of ribosome biogenesis. In yeast, the primary 35 S pre-rRNA assembles a number of 40 S subunit processing factors and the small subunit (SSU) processome containing the four known core proteins of U3 snoRNP complexes and forms the 90 S ribosomal precursor (6, 46, 47). hNop56p-associated pre-rRNPs contained human orthologs of all four core proteins of yeast U3 snoRNPs, namely hNop56p, fibrillarin, hNop5/Nop58p and a 15.5-kDa protein. These results suggest a role for hNop56p in pre-rRNA modifications during early stages of ribosome biogenesis, as is the case for yeast Nop56p. However, we did not identify any other transacting factors of the SSU processome or 40 S processing factors in the hNop56p-associated pre-rRNPs. Thus, hNop56p-associated pre-rRNPs differ from isolated yeast 90 S pre-ribosomes that contain most of the 28 components of the SSU processome and many 40 S processing factors (47). A variety of 60 S pre-ribosomes have been purified from yeast cells using a series of trans-acting factors as bait proteins. These pre-ribosomes are products of late stage ribosome biogenesis and contain a number of trans-acting factors that probably function as 60 S processing factors (11-13, 48). Thus, we compared the non-ribosomal proteins in the hNop56p-asso- a Protein localization in the nucleolus is adopted from the results as described (63,64). ciated pre-rRNPs with the yeast 60 S processing factors with respect to amino acid sequence similarity. Six proteins were assigned as potential human 60 S processing factors, namely hypothetical protein FLJ10377 (corresponding to yeast Nop4/ Nop77p), peter pan homolog (Ssf1p), EBNA1-binding protein 2 (Ebp2p), BRIX (Brx1p), DKFZP564M182 protein (Nsa3/Cic1p), and a putative nucleotide-binding protein (Nug1p). Fatica and Tollervey (6) proposed that at least four stable pre-rRNP complexes form during biogenesis of the yeast 60 S ribosomal subunit, and they classified the pre-60 S rRNP complexes into four categories: pre-60 S E 1 forms early during stage 1, pre-60 S E 2 during stage 2, pre-60 S M during the middle stage, and pre-60 S L forms during the late stage. By these criteria, hNop56passociated pre-rRNPs most likely represent pre-60 S E 1 because these complexes contain the largest number of transacting factors (Nop4/Nop77p, Ssf1p, Ebp2p, and Brx1p) found in common with the yeast pre-60 S E 1 purified using Ssf1p as bait (11).
hnRNP U is another potential trans-acting factor that we assigned to hNop56p-associated pre-rRNPs. This protein is thought to play a role in hnRNA processing, although its involvement in ribosome biogenesis has not yet been established in mammals. Its yeast ortholog, Nop3p, exhibits 37% sequence identity with human hnRNP U and binds RNA through its glycine-and arginine-rich domain as well as two RNA recognition motifs (49) (Supplemental Material Table II). Nop3p is essential for yeast ribosome biogenesis as its depletion results in a defect in 27 SB pre-rRNA, a component of pre-60 S complexes E 1 and E 2 that participates in 60 S subunit maturation (49). In addition, we identified the two RNA helicases DDX9 and DDX21 in hNop56p-associated pre-rRNPs as additional trans-acting factors possibly involved in ribosome biogenesis in mammalian cells. The yeast ortholog of DDX21, Dbp3p, is required for efficient conversion of the pre-rRNA 27 SA 2 to 27 SA 3 in pre-60 S E 1 (23). Thus, the human orthologs of Nop3p and Dbp3p support our proposal that hNop56p-associated pre-rRNPs represent the human equivalents of the yeast pre-60 S E 1 .
Other trans-acting factors found in hNop56p-associated pre-rRNPs include nucleolin and B23, both of which are unique to mammalian pre-rRNPs. Nucleolin participates in multiple stages of early pre-rRNA processing such as rDNA transcription, pre-rRNA processing, and assembly of ribosomal proteins and rRNAs (31)(32)(33)(34). The specific interaction of nucleolin with 5Ј-ETS sequences is required for the first step of pre-rRNA processing (33), which is consistent with the observation that the yeast counterpart Nsr1p is required for efficient pre-rRNA processing at sites A 0 to A 2 (50,51). Thus, it is well established that nucleolin and its yeast ortholog exhibit functional similarities in ribosome biogenesis. However, whereas nucleolin is found associated with hNop56p-and parvulin-associated pre-rRNPs isolated from mammalian cells (15), Nsr1p has never been identified in yeast pre-ribosomal complexes, suggesting that the mode of action of nucleolin in ribosome biogenesis differs from that of its yeast ortholog.
On the other hand, no yeast orthologs have been identified for another trans-acting factor, B23. This protein is an endoribonuclease that processes the ITS2 within the pre-rRNA, suggesting its involvement in pre-60 S complexes during the middle to late stages of 60 S biogenesis (35). Additionally, there are no reports of a yeast endoribonuclease that processes the ITS2. Rather, yeast remove the ITS2 using an enzyme complex composed of 11 exoribonucleases termed the "exosome" during the late stage of pre-60 S processing (52,53). However, we found no evidence of exosome components in hNop56p-associated pre-rRNPs. In total, the data indicate that hNop56p appears to function in pre-ribosomes during the early to middle stages of 60 S ribosome subunit maturation. However, the presence of the mammalian-specific trans-acting factor B23 and the unique association of nucleolin with the mammalian pre-rRNP complexes indicate that some of the fundamental processes of mammalian ribosome biogenesis differ from those of yeast.
A striking finding of this study is that, within the hNop56associated pre-rRNP, hNop56p interacts directly with treacle, the TCS gene product. The TCS disorder affects craniofacial development and has an incidence of approximately 1 in 50,000 live births (54,55). Whereas TCS is inherited as an autosomal dominant allele of the TCOF1 gene, 60% of the cases result from a de novo mutation (56). Almost all TCOF1 mutations reported to date have been either deletions, insertions, splicesite mutations, or nonsense mutations, indicating that TCS results from loss-of-function of the TCOF1 gene product (41,43,57,58). Furthermore, generation of Tcof1 heterozygous mice by targeted mutagenesis elicits the TCS phenotype. These observations suggest that treacle haploinsufficiency underlies this disorder. However, the normal function of treacle remains unclear.
Positional cloning identified TCOF1 (41) as a gene encoding the serine/alanine-rich protein, treacle (42,43). The treacle amino acid sequence includes several intriguing features. First, the protein localizes to the nucleolus via several nuclear localization signals within its C-terminal region (45,59). Second, it has a large central repeat domain containing numerous potential casein kinase II phosphorylation sites, indicating a possible regulatory mechanism. Third, treacle exhibits weak but significant similarity to the mouse nucleolar phosphoprotein Nopp140 (60) that is believed to act as a chaperone that mediates pre-rRNP export from the nucleus as well as ribosomal protein import from the cytoplasm during ribosome biogenesis. Because of the similarities between treacle and Nopp140, treacle may also function as an RNP chaperone (45). Thus, our finding that there is a direct interaction between hNop56p and treacle within hNop56p-associated pre-rRNP complexes provides the first experimental evidence supporting this new putative function for treacle.
The fact that hNop56p and treacle interact directly suggests that hNop56p may act as a scaffold that enables hNop56passociated pre-rRNPs to bind treacle and thereby export RNPs from the nucleolus. Interestingly, we also identified the nucleolar and coiled-body phosphoprotein 1 (NOLC-1), a human counterpart of Nopp140, as a component of the hNop56p-associated pre-rRNP. However, we were not able to prove that NOLC-1 binds directly to hNop56p using either MALDI-TOF/MS or LC-MS/MS of RNase-treated complexes. Although the possibility still remains that the repeated MS analyses simply failed to identify NOLC-1 in the RNase-treated complexes, our observation suggests that treacle and NOLC-1 (Nopp140) differ in their mode of action during ribosome biogenesis. We also note that, unlike NOLC-1 (Nopp140), treacle neither localizes to small RNP-rich Cajal bodies nor associates with snoRNPs (44) and that its amino acid sequence is unrelated to NOLC-1. Thus, although both of these proteins act as chaperones, their functions may be somewhat distinct. This assertion suggests that hNop56p-associated pre-rRNPs may utilize these two distinct chaperones at different stages during ribosome biogenesis.
Although we demonstrated the possible involvement of treacle in ribosome biogenesis and its direct association with hNop56p, it remains unclear as to whether the functions of hNop56p and treacle in ribosome biogenesis are related to the molecular mechanism underlying TCS. The analysis of Tcof1 heterozygous mice demonstrated that the correct dosage of treacle is essential for survival of cephalic neural crest cells that contribute significantly to formation of branchial arches. Hence, dysmorphogenesis of the lower face may be a consequence of defective neural crest cell migration from the hindbrain to the branchial arches during early embryogenesis (61,62). If the functions of hNop56p and treacle in ribosome biogenesis are directly related to TCS, then a possible underlying mechanism is that a sufficient supply of ribosomes is essential for normal development. Proteomic analysis of developmental profiles of hNop56p-and treacle-associated pre-rRNPs during TCS progression in model animals may suggest plausible mechanisms for the TCS disorder.
Apart from treacle, 29 non-ribosomal proteins of unknown function in ribosome biogenesis were identified in this study. Of these, 20 proteins reportedly localize to the nucleus and/or the nucleolus (63,64) (Table II and Supplemental Material  Table III). The subcellular localization of the remaining nine proteins has not been reported to date. Whereas the nucleolus constitutes the site of ribosome biogenesis, recent data indicate that the nucleolus may also play crucial roles in other cellular processes (65,66), particularly given that it associates with other subnuclear domains (67,68). Among the 29 non-ribosomal proteins, SFRS protein kinase 1, U2 small nuclear ribonucleoprotein auxiliary factor (65 kDa), CGI-74 protein, U2 small nuclear RNA auxiliary factor 1, and hnRNP A1 are categorized as proteins involved in mRNA processing and export. It has been suggested that spliceosomal small nucleolar RNP biogenesis and mRNA processing/export occur within the nucleolus (69,70). Four proteins that we identified, namely eukaryotic translation elongation factor 2, IGF-II mRNA-binding protein 1, eukaryotic translation elongation factor 1 ␣1, and Tu translation elongation factor (mitochondrial), are involved in protein translation. Recently, several lines of evidence suggested that translation mechanisms such as nonsense-mediated mRNA decay (71) may exist within the nucleus. Nonsense-mediated mRNA decay is thought to contribute to the mRNA quality control mechanism by reducing the level of mRNAs that contain nonsense mutations. Furthermore, functional ribosomes exist in a number of distinct intranuclear foci (72), suggesting that protein synthesis may occur in the nucleus. Signal recog-nition particle (SRP) biogenesis may also occur within the nucleolus (73). In fact, the 14-kDa SRP protein was identified in hNop56p-associated pre-rRNPs. Although the significance of SRPs in the nucleolus remains obscure, SRP assembly is believed to be coordinated with ribosome assembly within the nucleolus or nucleus as a mutual quality control mechanism that ultimately ensures the proper translation of each mRNA (74). We also identified nine proteins in hNop56p-associated pre-rRNPs that are classified as transcriptional regulator proteins (Table II and Supplemental Material Table III). These include MYB-binding protein (P160) 1a, interleukin enhancer binding factor 3, DNA topoisomerase I, a nuclear phosphoprotein similar to Saccharomyces cerevisiae PWP, nucleosome assembly protein 1-like factor 1, interleukin enhancer binding factor 2, H1 histone family member X, H1 histone family member 3, and H2B histone family member E. The functional relationship between the ribosome and transcriptional regulators is not presently understood. However, given that ribosomal components are found associated with the site of transcription (75), these proteins may be involved in the transcription of pre-ribosomal components. The remaining proteins identified in hNop56p-associated complexes are cytoskeletal proteins such as ␣-tubulin, ␤-tubulin, ␣-actin, and ␥-actin. Yeast Nop56p contains C-terminal KK(E/D) repeats, and similar motifs in human microtubule-associating proteins (76,77) mediate microtubule binding (78). It has been speculated that KK(E/D) repeats aid the efficient and accurate segregation of the pre-rRNA processing machinery during mitosis (16). hNop56p has no KK(E/D) repeats but rather contains highly basic repeats at its C terminus, suggesting a similar function for hNop56 during mitosis. The involvement of hNop56p-associated pre-rRNPs in the above pathways remains unproven, and these RNPs may act independently of pre-ribosomal complexes in some cases. Clearly, additional studies are needed to elucidate the mode of action of hNop56p in pre-rRNP complexes and delineate its role in the newly discovered functions of the nucleolus and the nucleus.