Analysis of the Ribosomal Protein S19 Interactome*S

Ribosomal protein S19 (RPS19) is a 16-kDa protein found mainly as a component of the ribosomal 40 S subunit. Its mutations are responsible for Diamond Blackfan anemia, a congenital disease characterized by defective erythroid progenitor maturation. Dysregulation of RPS19 has therefore been implicated in this defective erythropoiesis, although the link between them is still unclear. Two not mutually exclusive hypotheses have been proposed: altered protein synthesis and loss of unknown functions not directly connected with the structural role of RPS19 in the ribosome. A role in rRNA processing has been surmised for the yeast ortholog, whereas the extracellular RPS19 dimer has a monocyte chemotactic activity. Three proteins are known to interact with RPS19: FGF2, complement component 5 receptor 1, and a nucleolar protein called RPS19-binding protein. We have used a yeast two-hybrid approach to identify a fourth protein: the serine-threonine kinase PIM1. The present study describes our use of proteomics strategies to look for proteins interacting with RPS19 to determine its functions. Proteins were isolated by affinity purification with a GST-RPS19 recombinant protein and identified using LCMS/MS analysis coupled to bioinformatics tools. We identified 159 proteins from the following Gene Ontology categories: NTPases (ATPases and GTPases; five proteins), hydrolases/helicases (19 proteins), isomerases (two proteins), kinases (three proteins), splicing factors (five proteins), structural constituents of ribosome (29 proteins), transcription factors (11 proteins), transferases (five proteins), transporters (nine proteins), DNA/RNA-binding protein species (53 proteins), other (one dehydrogenase protein, one ligase protein, one peptidase protein, one receptor protein, and one translation elongation factor), and 13 proteins of still unknown function. Proteomics results were validated by affinity purification and Western blotting. These interactions were further confirmed by co-immunoprecipitation using a monoclonal RPS19 antibody. Many interactors are nucleolar proteins and thus are expected to take part in the RPS19 interactome; however, some proteins suggest additional functional roles for RPS19.

RPS19 1 is a structural component of the ribosomal 40 S subunit. It was considered to have only a structural role until its loss-of-function mutations were identified in patients with a rare hematological disease, Diamond-Blackfan anemia (DBA) (OMIM 105650) (1)(2)(3). DBA is characterized by defective erythroid progenitor maturation and is the first human disease due to mutations in a structural ribosomal protein. Dysregulation of RPS19 has thus been surmised as the cause of this defective erythropoiesis, although the link between them is still unclear. The finding that most RPS19 mutations suppress the expression of the allele has suggested that haploinsufficiency is the main cause of the defect (4,5). However, some patients carry missense mutations in the RPS19 gene. Deficient nucleolar localization may lead to abnormal ribosome incorporation and has been found for four missense mutants (6, 7) 2 ; this means that the disease mechanism may not be univocal.
RPS19 expression is increased during the intense proliferation at the start of erythropoiesis compared with the maturation of precursors at its close (8). Enhanced erythroid burstforming unit formation after overexpression of a wild type transgene in CD34ϩ bone marrow cells from DBA patients (9) and depressed in vitro erythropoiesis when RPS19 is knocked down (10) are other illustrations of its role.
Like other ribosomal proteins (RPs), RPS19 translocates from the cytoplasm to the nucleus where it participates in ribosome biogenesis. In yeast its absence is associated with abnormal rRNA cleavage and defective 40 S biogenesis (11,12). It has recently been suggested that defective erythropoiesis in DBA is due to the faulty protein synthesis particularly evident in progenitors whose RPS19 levels are lower than in other tissues (13,14).
We have used a yeast two-hybrid system to show that RPS19 binds PIM1, a ubiquitous serine-threonine kinase whose expression can be induced in erythropoietic cells by several growth factors, such as erythropoietin (15). We also showed that in human 293T cells PIM1 interacts with ribosomes and may be involved in translational control (15). A role in translational control of specific transcripts has been shown for other ribosomal proteins (i.e. RPL13 and RPL26) (16,17).
It thus appears that RPS19, in addition to its structural role in the ribosome, is involved in ribosome biogenesis, specifically in rRNA processing and possibly in translation. These functions are probably assisted by interaction with different protein substrates.
In the study now reported, we used functional proteomics procedures to look for proteins interacting with RPS19 (18) and thus secure additional information regarding its function and regulation. We identified 159 RPS19-associated proteins. These included many ribosomal proteins and proteins with a known role in ribosome biogenesis. Furthermore the identification of proteins with other functions, such as translational control and splicing, indicates that RPS19 may also be involved in RNA processing/metabolism and translational control.
Expression and Purification of Fusion Proteins-The human RPS19 cDNA was amplified by RT-PCR (15) and cloned into pGEX-4T-1 (Amersham Biosciences) to generate plasmid pGEX-RPS19. As a further control we used a pGST-NTGATA1 construct that encodes for a GST fusion protein with the N-terminal domain of the human GATA1 transcription factor (19).
GST, GST-RPS19, and GST-GATA fusion proteins were expressed in Escherichia coli cells, strain BL21, by induction with 0.5 mM isopropyl 1-thio-␤-D-galactopyranoside for 1 h at 37°C. Bacteria were resuspended in PBS containing 1% Triton X-100, 1 mM EDTA, 1 mM PMSF, 1 g/ml leupeptin, 1 g/ml aprotinin, and 1 g/ml pepstatin. Bacterial extracts were sonicated and centrifuged to remove cell debris. GST proteins were purified by affinity binding to GST⅐Bind TM resin (Novagen, Madison, WI). Protein samples were separated by SDS-PAGE and compared with known concentrations of bovine serum albumin after Coomassie Brilliant Blue staining.
Affinity Purification-Whole cell extract was preincubated with GST resin (300 g of recombinant protein) for 1 h at 4°C. Unbound proteins were then incubated with the same quantity and volume of GST-RPS19 resin overnight at 4°C on a rocker. The resin was extensively washed with TM 0.1 buffer (0.1 M KCl in TM 0.0 buffer), and bound proteins were eluted with TM 0.5 buffer (0.5 M KCl in TM 0.0 buffer) and precipitated with 20% trichloroacetic acid. The pellets were washed twice with acetone, dried, and used for mass spectrometry. This experiment was repeated six times to provide enough samples.
Monoclonal Antibody against RPS19 -Immunization and screening for putative monoclonal antibodies have been carried out according to Cianfriglia et al. (20). Briefly BALB/c mice (age, 12 weeks) were repeatedly intraperitoneally injected (five times) with 30 g of purified GST-human RPS19 (the first injection was diluted with Freund's complete adjuvant; the second injection, after 10 days, was diluted with Freund's incomplete adjuvant; the other boosters, every 4 days, were with saline solution). Hybrid cells were obtained by fusion of myeloma cells (SP2/0-AG-14) with polyethylene glycol (Sigma) and were screened by ELISA with recombinant GST-RPS19. Positive clones were expanded, and the supernatant was analyzed by Western blotting. Highly positive hybridomas were cloned by limiting dilution, and the stable line C3 was selected for the production of antibody specific for RPS19. The heavy chain isotype of C3 monoclonal antibody is IgG1 with light chains as determined by a mouse hybridoma subtyping kit (Roche Applied Science).
Validation by Western Blot and Co-immunoprecipitation-To validate the MS/MS results, new preparations of GST-RPS19 pulldowns were subjected to Western blot analysis. Antibodies specific for PIM1 (Upstate, Charlottesville, VA), insulin-like growth factor 2-binding protein 1 (IGF2BP1) (IMP1), minichromosome maintenance-deficient protein 6 (MCM6), DDX5, and nucleolin (NCL) (C23) (Santa Cruz Biotechnology, Santa Cruz, CA) were used according to the manufacturer's instructions. Monoclonal anti-STAU1 antibody was a gift from Dr. Luc DesGroseillers (21) (University of Montreal, Montreal, Canada) and used at a dilution of 1:1000. The polyclonal antibody against DKC1 was a gift from Philip Mason (Washington University, St. Louis, MO) (22) and used at a dilution of 1:5000. All immunoblot detections were carried out using horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence (Amersham Biosciences) with the exception of the nucleolin blots where an alkaline phosphatase-conjugated secondary antibody was used (Sigma).
For co-immunoprecipitation analyses, 0.5% Triton X-100 was added to K562 whole cell extracts, prepared as described above. Extracts were precleared by incubation with protein G-agarose (Sigma) on a rocker for 1 h at 4°C.
The supernatant was incubated with an anti-RPS19 monoclonal antibody (hybridoma supernatant) and with protein G-agarose on a rocker at 4°C for 16 h. As a negative control, we used an antihemagglutinin monoclonal antibody (Santa Cruz Biotechnology).
SDS-PAGE, In-gel Digestion, Peptide Mapping, and Mass Spectrometry-The six pellets obtained by affinity purification were resuspended in SDS-PAGE sample buffer and pooled for one-dimensional electrophoresis. The total volume for each sample (GST and GST-RPS19) was 50 l. The two protein mixtures were fractionated by 8 -18% SDS-PAGE. Molecular masses of protein bands were esti-mated by using Precision Plus All Blue protein standards (Bio-Rad). Protein electrophoretic patterns were then visualized using GelCode Blue Stain Reagent (Pierce).
The GST-RPS19 and GST gel lanes were cut to create 65 2-mm slices per lane. Each slice was crushed and washed first with acetonitrile and then with 0.1 M ammonium bicarbonate. Protein samples were reduced by incubation in 10 mM dithiothreitol for 45 min at 56°C and alkylated with 55 mM iodoacetamide in 0.1 M ammonium bicarbonate for 30 min at room temperature in the dark as described previously (23). The gel particles were then washed with 0.1 M ammonium bicarbonate and acetonitrile. Enzymatic digestions were carried out with modified trypsin (Sigma) (10 ng/l) in 50 mM ammonium bicarbonate, pH 8.5, at 4°C for 45 min. The enzymatic solution was then removed. A new aliquot of the buffer solution was added to the gel particles and incubated at 37°C for 18 h. A minimum reaction volume sufficient for complete rehydration of the gel was used. Peptides were extracted by washing the gel particles in acetonitrile at 37°C for 15 min and lyophilized.
The peptide extract volumes were divided in two to inject the peptide mixtures two times. The analysis were performed by LCMS/MS with a Q-TOF hybrid mass spectrometer (Waters, Milford, MA) equipped with a Z-spray source and coupled on line with a capLC chromatography system (Waters) or alternatively by using the LC/MSD Trap XCT Ultra (Agilent Technologies, Palo Alto, CA) equipped with a 1100 HPLC system and a chip cube (Agilent Technologies). After loading, the peptide mixture (7 l in 0.5% TFA) was first concentrated and washed (i) at 1 l/min onto a C 18 reverse-phase precolumn (Waters) or (ii) at 4 l/min in a 40-nl enrichment column (Agilent Technologies chip) with 0.1% formic acid as the eluent. The sample was then fractionated on a C 18 reverse-phase capillary column (75 m ϫ 20 cm in the Waters system, 75 m ϫ 43 mm in the Agilent Technologies chip) at a flow rate of 200 nl/min with a linear gradient of eluent B (0.1% formic acid in acetonitrile) in A (0.1% formic acid) from 5 to 60% in 50 min. Elution was monitored on the mass spectrometers without any splitting device. Peptide analysis was performed using data-dependent acquisition of one MS scan (mass range from 400 to 2000 m/z) followed by MS/MS scans of the three most abundant ions in each MS scan. Dynamic exclusion was used to acquire a more complete survey of the peptides by automatic recognition and temporary exclusion (2 min) of ions from which definitive mass spectral data had been acquired previously. Moreover a permanent exclusion list of the most frequent peptide contaminants (keratins and trypsin doubly and triply charged peptides: 403. 20 Data Analysis-Raw data from LCMS/MS analyses were converted into a Mascot format text to identify proteins by means of the Matrix Science software (24). The protein search was governed by the following parameters: non-redundant protein sequence database (NCBInr, January 24, 2006 download, 3,229,765 sequences), specificity of the proteolytic enzyme used for the hydrolysis (trypsin), taxonomic category of the sample (Homo sapiens), no protein molecular weight was considered, up to one missed cleavage, cysteines as S-carbamidomethylcysteines, unmodified N-and C-terminal ends, methionines both unmodified and oxidized, putative pyro-Glu formation by Gln, precursor peptide maximum mass tolerance of 150 ppm, and a maximum fragment mass tolerance of 300 ppm. In the experience of the authors' laboratory all the MS/MS spectra displaying a Mascot score (24) higher than 38 show a good signal/noise ratio leading to an unambiguous interpretation of the data. Individual MS/MS spectra for peptides with a Mascot score (24) equal to 38 were inspected manually and only included in the statistics if a series of at least four continuous y or b ions were observed.
In Silico Analysis-A list of primary (direct) and secondary (indirect) protein-protein interactions of RPS19 was created using the webavailable Human Protein Reference Database (www.hprd.org). In April 2006, the database contained 20,097 human protein entries, 33,710 documented protein-protein interactions, and 171,677 links to the PubMed literature. Primary interactions of RPS19 were screened for protein interactors to define an in silico interaction map with the indirect protein partners. This map was then compared with the RPS19 protein partners identified in this study. In addition a list of primary interactions was created by HPRD for each identified protein.
Lastly the list of RPS19 protein partners was compared with the Nucleolar Proteome Database (www.lamondlab.com/NoPDB) (25) and with the Pre-Ribosomal Network yeast database (www.pre-ribosome.de/Home.html) (26). Ortholog Saccharomyces cerevisiae gene names were determined using the web-available database Ensembl (www.ensembl.org).

Identification of RPS19-interacting Proteins-
To determine the RPS19 interactome in K562 cells, we performed LCMS/MS analysis of proteins purified by pulldown experiments on a GST-RPS19 affinity resin. The proteins eluted from the GST-RPS19 or the negative control GST resins were fractionated by 13-cm 8 -18% SDS-PAGE and revealed by colloidal Coomassie stain (Fig. 1). SDS-PAGE indicated that the RPS19-associated proteins span a broad molecular weight range. The two major bands at 26 kDa (Fig. 1 pulldown experiments, aliquots of the proteins eluted from the GST or the GST-RPS19 resins were analyzed by Western blotting using an anti-PIM1 antibody. PIM1 (i.e. the positive control) was identified in the GST-RPS19 lanes only (see Fig. 4).
The procedure described under "Experimental Procedures" gave 65 peptide mixture samples from each lane. The peptide extract volumes were divided in two to analyze peptide mixtures two times by LCMS/MS. These duplicates showed a high level of reproducibility where in all cases identifications from the first analysis were confirmed from the second one. Peptide mixtures deriving from the GST lane constituted our control for the analysis of GST-RPS19 lane and were therefore always injected before the peptides from the GST-RPS19 lane. Mass spectrometry data were then analyzed with the Mascot software on the NCBI human protein sequence database. To select proteins that interact specifically with RPS19, we subtracted species common to the GST and GST-RPS19 lanes (Fig. 1). These proteins are shown in Supplemental Table 1. Table I displays the complete list of RPS19 protein interactors identified in this study. Proteins are grouped according to their known function, and for each identification the human gene name, the corresponding protein name, and the ortholog S. cerevisiae gene name is reported. Supplemental Table 2 reports for each protein entry the identified peptides together with their sequences, the observed mass errors on the precursor peptides, the Mascot score for each peptide, and the protein sequence coverage expressed as the number of amino acids spanned by the identified peptides divided by the sequence length. All protein species identified by a single peptide were further checked. First the peptide sequence stretch, manually verified, was searched on the Basic Local Alignment Search Tool (BLAST) software at the NCBI web site (ncbi.nlm.nih.gov/blast) against human taxonomy. When other matches were possible, the candidate was removed from the list. The remaining single peptide protein species were added to the list only when involved in protein complexes known to interact with mRNA/rRNA or reported to interact with one of the proteins identified in this study (27)(28)(29)(30)(31)(32). Fig. 2 (A-D) shows the MS full scan, the MS/MS scan, and the amino acid sequence relative to four identified RPS19associated proteins: IGF2BP1 ( Fig. 2A), MCM6 (Fig. 2B), DDX5 (Fig. 2C), and STAU1 (Fig. 2D). STAU1 is an example of protein species identified by a single peptide. Supplemental Fig. S3 shows additional examples like CCT2 (A), DDX17 (B), and NOLA3 (C).
The 159 human proteins identified in this study were divided into Gene Ontology functional groups as shown in Fig.  3A: NTPases (ATPases and GTPases, five proteins), hydrolases/helicases (19 proteins), isomerases (two proteins), kinases (three proteins), splicing factors (five proteins), structural constituents of ribosome (29 proteins), transcription factors (11 proteins), transferases (five proteins), transporters (nine proteins), DNA/RNA-binding protein species (53 pro-teins), other (one dehydrogenase protein, one ligase protein, one peptidase protein, one receptor protein, and one elongation factor), and 13 proteins of still unknown function. They were also grouped according to their Gene Ontology cellular localization (Fig. 3B) and the biological processes in which they are involved (Fig. 3C). Moreover according to the Human Nucleolar Database, 101 are nucleolar proteins (Supplemental Table 3); many of these are part of the 90 S preribosome as assessed by comparison with the yeast Pre-Ribosomal Network (Supplemental Table 4).
Among the interactors we identified RPS8, a ribosomal protein found in a previous yeast two-hybrid study. 3 In the same study we also revealed PIM1 (15), which was not detected in the proteomics analysis despite its presence in the eluate from GST-RPS19 resin (Fig. 4) and in the immunoprecipitate obtained with the anti-RPS19 monoclonal antibody (Fig. 5). The difference is presumably ascribable to the sensitivity limitations of proteomics analysis compared with antibody-based assays.
Validation of MS/MS Data by Western Blotting and Immunoprecipitation-To corroborate the authenticity of the proteins identified by MS/MS, we confirmed the presence of representative proteins in the eluates from GST-RPS19 resins by immunoblotting: the serine-threonine kinase PIM1, IGF2BP1, MCM6, the DEAD box polypeptide 5 (DDX5), Staufen (STAU1), dyskerin (DKC1), and NCL (Fig. 4).
Affinity purification was also performed using as negative controls a GST-GATA1 protein and different amounts of the GST protein. The bound proteins were eluted from the resins and analyzed by Western blot using specific antibodies for three selected interactors (DDX5, DKC1, and NCL). All these negative controls confirmed the specificity of the interaction between RPS19 and the proteins analyzed (Supplemental Fig.  S1).
In addition, we performed immunoprecipitations using K562 lysates and a monoclonal antibody to RPS19 to show that the same interactions occur in living cells. Immunoprecipitated proteins were resolved by SDS-PAGE and analyzed by Western blotting with specific antibodies (Fig. 5 and Supplemental Fig. S2). In Fig. 5 we omitted the results regarding the positive co-precipitation of STAU1 because the close co-migration of the immunoglobulin heavy chains affects the quality of the data (as shown in Supplemental Fig. 2A). Supplemental Fig. S2, A, B, and C, shows the whole image of the co-immunoprecipitation assay.
The already known primary interactors of RPS19 reported in the in silico analysis were not identified in this proteomics study. However, they were identified interacting with RPS19 in very particular conditions. In the case where FGF2 was reported to interact with free RPS19, in fact the GST pulldown experiment was performed with only the cytoplasmic extracts of NIH3T3 or ECV304 cells (33). In the case of complement component 5 receptor, the protein-protein interaction was reported in extracts of a rheumatoid arthritis synovial lesion when a covalent dimer of RPS19 by transglutamination occurs (34). In the case of RPS19-binding protein the specific antibody is not available (35) to perform an antibody-based assay to complement mass spectrometric data.
Our analysis shows that several proteins identified in this study interact with each other. DDX5, PES1, DDX21, GTPBP4, NOL5A, and NCL, for example, interact with RPL4, RPL6, RPL7a, RPL10a, RPLP0, and RPS3. Their relationship with RPS19, however, is not illustrated in the HPRD database nor in the literature, and they are therefore new RPS19 partners.

DISCUSSION
This study represents the first global, high throughput functional proteomics approach to identify the proteins that interact with RPS19. Our analysis of the GST-RPS19 pulldown revealed 159 proteins, most of them not previously known to associate with RPS19. On the other hand, in silico analysis and PubMed data show that many proteins interact with each other. They may thus participate with RPS19 in the same multiprotein complex or complexes.
It is known that ribosomal proteins are involved at different stages of ribosome biogenesis and/or in distinct translation steps (36). In particular, they have been thought to play a central role in rRNA processing, protein assembly, RNA folding, transport of the ribosomal precursors, stabilization of the subunit structure, and/or interaction with other factors required for either ribosome biogenesis or translation (37)(38)(39). Their involvement in cotranslational processes, such as the interaction with protein folding factors at the exit tunnel of the Proteins from K562 whole cell extracts were affinity-purified with GST and GST-RPS19 resins. Total lysates (K562 lysate) and bound proteins eluted from the GST control resin (GST) or the GST-RPS19 resin (GST-RPS19) were analyzed by Western blotting with antibodies specific to the indicated proteins. All blots were revealed by the chemiluminescence method except for NCL, which was revealed by alkaline phosphatase. ribosome (40,41), cotranslational translocation (42,43), and important enzymatic activities for ribosome function, e.g. the mRNA helicase activity of bacterial ribosomes (44), has also been proposed.
Interestingly most proteins reported in this study, such as nucleolar or ribosomal proteins, play a role in processes related to RPS19. It should be stressed that we used a total cell lysate and not a nuclear extract and that the complex formation was extracellular. Nevertheless proteins abundant in cytoplasm were not found. This suggests that the structure of the recombinant RPS19 protein is functionally suitable to recruit multiple cellular partners.
The identification of RPs belonging to the small and the large subunits suggests that we have purified components of the preribosome (90 S), the structure formed before processing of the pre-rRNA. The subsequent cut at a specific site (A2) divides these subunits. The preribosome is a highly dynamic structure that comprises more than 150 non-ribosomal proteins with various activities, including nucleases, RNA helicases, GTPases, AAA ATPases, kinases, etc. (for reviews, see Refs. 45 and 46). We have, indeed, found 23 of 31 proteins with orthologs in the yeast preribosome network that belong to the 90 S subunit. Many interactors are shared between FIG. 5. Co-immunoprecipitation. Proteins from K562 cell lysates were immunoprecipitated with a monoclonal anti-RPS19 or an antihemagglutinin antibody as negative control. Immunocomplexes were fractionated by SDS-PAGE, blotted on nitrocellulose, and revealed by the specific antibodies. All blots were revealed by the chemiluminescence method except for NCL, which was revealed by alkaline phosphatase. IP, immunoprecipitate; HA, hemagglutinin.

Analysis of RPS19-interacting Proteins
RPS19 and parvulin, a peptidyl-prolyl isomerase involved in early ribosome biogenesis (i.e. L3, L4, L6, L7, L7a, L8, L10a, L14, S3, S4X, S6, S8, and DDX18) (47). The interaction with most of the RPs essential for the transport of the small subunit from the nucleus to the cytoplasm (i.e. RPS10, RPS26, RPS3, and RPS2) and to the exportin XPO (known to control the 40 S and 60 S export) suggests a role for RPS19 in this process. This is in agreement with two recent reports of its involvement in the early processing of rRNA (11) and possibly in its export from the nucleus (12). Concordantly the greater portion of the RPS19-interacting proteins identified in this study includes proteins involved in pre-rRNA processing, such as RNA helicases, and major components of the box C-D small nucleolar RNAs (48,49), such as fibrillarin and Nop56.
We also found major components of the H/ACA box small nucleolar RNP complex, i.e. dyskerin, NOLA1, and NOLA3. This complex (that includes a fourth protein, NOLA2) is required for the site-specific pseudouridylation of rRNA involved in the early stages of ribosome biogenesis (50). Both 18 S rRNA production and rRNA pseudouridylation are impaired if any one of the four proteins is depleted.
A further group of interactors includes proteins controlling protein synthesis, such as proteins involved in cotranslational translocation (42,43) (such as signal recognition particle 68) and translation regulators, such as IGF2BP1 and STAU1. Other ribosomal proteins (i.e. RPL13 and RPL26) (16,17) are known to regulate translation of specific transcripts. It is intriguing that RPS19 could have a similar role. Our previous studies showing interactions of PIM1 and RPS19 on the 40 S subunit suggested such a role (15).
Lastly this study identified proteins with more diverse cellular functions. These included proteins such as integrins, proteasome components, and kinases. Further studies are needed to clarify their involvement in the RPS19 interactome.
The scenario disclosed by our study clearly shows that RPS19 is definitely involved in RNA processing and metabolism and perhaps in translation control. Although it must be stressed that our results do not take the spatial-temporal dimension of RPS19 interactome into account, future experiments will be directed toward the comprehension of this point.
It is intriguing that among the direct or indirect RPS19 interactors we also found proteins involved in pathologies with phenotypes similar to DBA (14). These include the following: 1) DKC1, responsible for dyskeratosis congenita (OMIM 305000), that shares bone marrow failure with DBA; 2) RPL24, whose spontaneous defect in mice produces growth retardation and skeletal malformations (51); 3) TCOF1, responsible for the Treacher-Collins syndrome (OMIM 154500), which shares some malformations with DBA, and that interacts with NOL5A and UBTF; and 5) SBDS (OMIM 260400), responsible for the Schwachman-Diamond syndrome, that interacts with nucleolin. This suggests a link between the ribosomal diseases, possibly a common pathogenetic mechanism.
In short, we have identified several new protein interactions with RPS19. This should lead to a fuller understanding of its activities and a more complete picture of its cellular roles and/or regulation. A clearer understanding of the function of RPS19 could help to elucidate the pathogenesis of DBA.