Ribosomes are responsible for protein synthesis in all cells. Ribosomal protein S19 (RPS19) is one of the 79 ribosomal proteins (RPs) in vertebrates. Heterozygous mutations in RPS19 have been identified in 25% of patients with Diamond-Blackfan anemia (DBA), but the relationship between RPS19 mutations and the pure red-cell aplasia of DBA is unclear. In this study, we developed an RPS19-deficient zebrafish by knocking down rps19 using a Morpholino antisense oligo. The RPS19-deficient animals showed a dramatic decrease in blood cells as well as deformities in the head and tail regions at early developmental stages. These phenotypes were rescued by injection of zebrafish rps19 mRNA, but not by injection of rps19 mRNAs with mutations that have been identified in DBA patients. Our results indicate that rps19 is essential for hematopoietic differentiation during early embryogenesis. The effects were specific to rps19, but knocking down the genes for three other RPs, rpl35, rpl35a and rplp2, produced similar phenotypes, suggesting that these genes might have a common function in zebrafish erythropoiesis. The RPS19-deficient zebrafish will provide a valuable tool for investigating the molecular mechanisms of DBA development in humans.
Diamond-Blackfan anemia (DBA) is characterized by diminished numbers of erythroid progenitors in bone marrow (BM) in early infancy. In some cases, patients also show diverse physical abnormalities, such as upper limb malformations, short stature and kidney dysfunction (1,2). DNA analysis of a DBA patient with a translocation at 19q13 identified ribosomal protein (RP) S19 as a candidate disease gene for DBA (3). Subsequently, heterozygous mutations in RPS19 were detected in 25% of 172 patients (4). RPS19 is one of 79 RPs, and is expressed in every cell where protein synthesis occurs. It is still unclear how the mutations in such a ubiquitously expressed gene specifically affect erythropoiesis.
Several attempts have been made to address this question, mainly using cell lines. For instance, cells with siRNA-mediated knockdown of the RPS19 gene proliferate less (5) and have defects in erythroid differentiation similar to those seen in DBA patients (6). Apoptotic changes in progenitor cells have been suggested to be a major factor for anemia development, as evidenced by increased apoptosis in CD34+ cells from DBA patients as well as in siRNA-treated RPS19-deficient cells (7). Others have suggested that cell cycle arrest at G0/G1, not apoptosis, is more critical for the development of anemia (8). Regardless of the mechanism involved, the question remains how a lack of RPS19 protein disturbs the proliferation of erythroid precursors.
Recently, defects in ribosome biogenesis have been proposed to be important in DBA pathogenesis. RPS19 mutations result in the accumulation of premature ribosomal RNAs (rRNAs) in patient-derived cells and in RPS19-deficient cell lines (9,10). When transfected into human cell lines, RPS19 proteins with the mutations seen in DBA patients fail to associate with ribosomes (11). Moreover, mutations in two other RP genes, RPS24 and RPS17, have also been identified in a small but significant number of DBA patients (12,13), suggesting that changes in ribosomal function could be responsible for the defective erythropoiesis in DBA.
In vertebrates, ribosomes are composed of four rRNA species and 79 different proteins (14–16). Although RPs are essential for the assembly of ribosomes, not much is known about their role, if any, during translation. Therefore, elucidating the currently unknown functions of RPs should provide significant insight into the pathogenesis of DBA and other ribosome-associated diseases. To date, only a single animal model, Rps19-knockout mice, has been developed to explore the role of RPS19 in DBA (17). However, Rps19-heterozygous mice in this model system display no abnormalities in any organ, including the hematopoietic system, whereas the null mice die embryonically. Here, we report the development of rps19-knockdown zebrafish that show marked reduction in erythrocyte numbers during embryogenesis.
Phenotypes of rps19-knockdown embryos
Mutations in universally expressed genes, such as RP genes, are generally assumed to result in systemic abnormalities. However, in our previous study, embryos injected with Morpholino antisense oligos (MOs) against 20 different RP genes showed specific phenotypes depending on the RP gene targeted (18). This suggests that each RP has a specific, although unknown, function during early development in zebrafish. To investigate the specific role of the RPS19 gene in erythropoiesis, we knocked down the zebrafish ortholog (rps19) using MOs and analyzed the effect on the synthesis of blood cells.
The coding region of rps19 shares 78% nucleotide and 88% amino acid identity with its human ortholog. Although gene duplication is common in zebrafish, available information from public databases suggests that rps19 exists as a single copy in the genome. We targeted this gene using an MO aimed at its translation initiation site. Following injection of this MO at the one-cell stage, we compared the morphological features of the morphants with wild-type siblings at about 24 h postfertilization (hpf). At this stage, pigmentation in the retina begins to appear and somitogenesis is completed. The rps19 morphants showed incomplete brain subdivisions and a ventrally bent tail (Fig. 1). We also observed other phenotypes commonly associated with RP morphants (18), such as an incomplete yolk sac extension and a rough surface appearance. The embryos injected with a control MO did not display any morphological changes. All the rps19 morphants died by 10 days postfertilization.
Reduced red blood cell synthesis in rps19-knockdown embryos
Blood circulation in wild-type zebrafish is easily visible in the posterior cardinal vein and the common cardial vein by 26 hpf. However, the process was delayed by 2–3 h in the rps19 morphants. Therefore, we compared the circulation pattern of the morphants at 29 hpf with that of wild-type embryos at 26 hpf. Rps19 morphants showed an obvious reduction in the circulating blood cells compared with the control embryos (Supplementary Material, Fig. S1–S4), and this reduction continued even at later stages of development.
When observed at 48 hpf, the heart in normal embryos appeared red-colored due to a high density of blood cells; however, it was almost transparent in rps19 morphants, indicating a decreased number of blood cells (Fig. 2B, arrow). To further confirm this reduction, we performed hemoglobin staining on the embryos. As expected, the hemoglobin staining observed in the cardial vein (Fig. 2C, light gray dots) was markedly decreased in the morphants. To investigate whether this reduction in blood cells was common in RP deficient embryos, we carried out hemoglobin staining of 19 additional RP morphants. All the 19 RP morphants showed varying degrees of morphological abnormalities; however, there was no correlation between the staining intensity and severity of the associated morphological defects (described later). This suggests that defective blood cell production is a consequence of a specific RP deficiency.
Primitive erythropoiesis defects in rps19 morphants
To determine the onset of defective erythropoiesis, we analyzed the expression pattern of gata1, an early erythroid gene, by whole mount in situ hybridization. Because the initiation of circulation was slow in the morphants due to the developmental delay, we examined gata1 expression at different developmental stages (23–27 hpf) in wild-type and MO-injected embryos. The gata1 expression in the intermediate cell mass (ICM) peaked around 24 hpf in the wild-type embryos, whereas it peaked around 26 hpf in the morphants (Fig. 3A, b and e). However, when compared to gata1 mRNA levels at appropriate developmental stages, the expression level of gata1 mRNA was not significantly different between the wild-type embryos and the morphants until the onset of circulation (Fig. 3A, a and d, b and e). In contrast, when erythroblasts matured into erythrocytes during circulation (19), gata1 mRNA was detected in the yolk region of wild-type embryos, but not in the morphants, although their gata1 expression patterns at ICM were similar (Fig. 3A, c and f).
To further confirm the specificity of the erythroid defects caused by rps19 deficiency, we examined the expression of vascular and myeloid lineage genes. The expression of tie-1 (Fig. 3B) and fli-1a (data not shown) during vasculogenesis was similar in the wild-type and MO-injected embryos. Similarly, semi-quantitative RT–PCR of other myeloid genes, including pu.1, l-plastin and mpo, did not show any significant difference in the transcript levels at 24 hpf (Fig. 3C). To confirm the localization of these genes, we carried out in situ hybridization and found no obvious differences in the expression patterns between the wild-type embryos and the morphants (Supplementary Material, Fig. S5). Even at 48 hpf, a stage when significant reduction of erythrocytes occurred in the morphants as judged by hemoglobin staining (Fig. 2D), the expression level of l-plastin and mpo was similar in both types of embryos (data not shown). These data indicate that maturation and proliferation of the erythroid lineage is the main defect of the morphants.
Phenotypic rescue in the morphants by rps19 mRNA synthesized in vitro
Since the reduced blood cell phenotype was specific to RPS19 deficiency, we examined whether a synthesized rps19 mRNA could rescue this phenotype in embryos. For this, we incorporated a five-base change into the MO-target site of the rps19 mRNA to prevent in vivo binding of the MO to the synthesized mRNA (Fig. 4A). Co-injection of this mRNA (1.0 µM) with the rps19 MO (60 µM) resulted in almost complete recovery of blood cells, as observed by hemoglobin staining and observation of gross morphology (Fig. 4B). These results confirm that the decreased blood cell synthesis is directly related to the RPS19 deficiency in zebrafish.
Several types of mutations, including allelic loss, point mutations, insertions and deletions, have been identified in the RPS19 gene of DBA patients (4,20). Among these, a missense mutation at the 62nd amino acid and a nonsense mutation at the 94th amino acid, both in exon 4 of the gene, occur at high frequency. We assessed whether mRNAs containing these patient-type mutations could rescue the rps19 phenotype (Fig. 4A). However, when we injected these two mutated mRNAs at a 1.0 µM concentration, which is sufficient for rescue by non-mutated mRNA, the embryos displayed more severe developmental defects than the rps19 morphants (data not shown). Even in wild-type embryos, injection of these mRNAs at the same concentration resulted in morphological abnormalities (data not shown). Injection of these mRNAs at a lower concentration (0.5 µM) did not induce any additional deformities, but also did not rescue the blood cell production and other associated phenotypes of the knockdown (Fig. 4C). These results indicate that the rps19 mutations seen in patients remove the normal function of the RPS19 protein and therefore render them unable to rescue red blood cell synthesis in the zebrafish displaying defective erythropoiesis.
Reduced red blood cell synthesis in other RP morphants
Recently, it was proposed that the pathogenesis of DBA might be linked to functional changes in the translational machinery due to defective RPs (21,22). We, therefore, investigated the possibility that RP genes other than RPS19 could contribute to the onset of DBA.
We examined the circulation pattern and the red blood cell density in 20 different RP morphants at various stages of development (26–48 hpf) by using live video imaging and hemoglobin staining. Most of the RP morphants displayed an initial delay in circulation compared to the wild-type embryos, which may be due to a general effect of RP deficiency. Therefore, we focused on whether these initial circulatory defects recovered during later stages of development. We categorized the defects into three different groups: severe, moderate and almost normal. Our observations revealed that the circulatory defects and blood cell reduction were consistent in three RP morphants in addition to rps19: rpl35, rpl35a and rplp2 (Table 1). Although the morphological phenotypes differed among them, the blood cell defects were similar to those of the rps19 morphants, as indicated by a dramatic decrease in hemoglobin staining (Fig. 5). Thus, these four RP genes might have a common function in zebrafish erythropoiesis.
|Phenotypes||Blood cell circulation||Blood cell density|
|Posterior caudal vein||Cardial vein||Hb staining|
|Phenotypes||Blood cell circulation||Blood cell density|
|Posterior caudal vein||Cardial vein||Hb staining|
RPS19-deficient zebrafish as a DBA model
Since the first report of DBA in 1938 (23), not much has been learned about its pathogenic mechanisms. Although the RPS19 gene was identified as a candidate disease gene in 1999 (3), we still have not been able to understand this disease completely. Recent studies of DBA and other inherited BM-related diseases indicate a possible connection between the translational machinery and BM defects (24,25). For instance, the SBDS gene, predicted to have a role in ribosome biogenesis, is mutated in Shwachman-Diamond Syndrome, a disease characterized by exocrine pancreatic insufficiency, BM failure and other somatic abnormalities (26,27). Similarly, in dyskeratosis congenita and cartilage-hair hypoplasia, mutations have been identified in the genes that encode proteins involved in ribosome biogenesis (28,29). However, like DBA, there is currently no direct evidence to prove the relationship between the mutations in these genes and the resultant clinical features of the diseases. Although the molecular mechanisms of pathogenesis are not understood in these BM disorders, the ribosome seems to be a common link between them. Therefore, elucidating the underlying mechanism in DBA may bring new insights into other ribosome-associated diseases as well.
Studies using cell lines and cells derived from DBA patients demonstrated that changes in basic cellular functions, such as apoptosis and cell cycle arrest, contribute to DBA development (7,8). However, these observations, although important, focused on the erythroid lineages, and changes in other tissues may have been overlooked. To investigate the mechanisms underlying the erythroid specificity of RPS19 depletion, studies in animal models are necessary. A previous attempt to use Rps19-knockout mice was unsuccessful because the complete loss of the gene resulted in embryonic lethality, and loss of a single allele did not produce any unusual phenotype (17,30). Although it is not clear why the Rps19 heterozygous mice developed normally, given that DBA patients have heterozygous mutations, it seems that Rps19 knockout mice are not a suitable animal model for DBA. As an alternative approach, we knocked down rps19 in zebrafish. In this study, we successfully demonstrated that the loss of rps19 function in zebrafish recapitulates the DBA phenotype of a severe decrease in the production of erythroid cells. However, early erythropoiesis was not disrupted, as suggested by the normal expression of gata1 (Fig. 3). In contrast, the level of mature erythrocytes was significantly lower in the morphants. This may indicate the existence of an unknown mechanism causing pure red-cell anemia. There are several advantages to using zebrafish that make this model appropriate for studying the molecular mechanism of DBA. First, the morphology and the maturation process of erythroid cells closely resemble those in mammals. Second, the blood cells in the early stages of development consist mainly of erythroid progenitors, which are the primary targets in DBA, making the analysis of the hematopoietic system easier. Third, the embryos can survive for several days even in the absence of blood cells, and hence are suitable for anemia research (31).
Effects of mutated S19 proteins on embryogenesis
In this zebrafish system, the blood cell reduction was easily rescued by rps19 mRNA, but not by mRNAs with DBA patient-type mutations. Interestingly, these mutated mRNAs, when overexpressed, led to severe phenotypes different from those seen in rps19 morphants. A previous in vitro study demonstrated that mutated S19 proteins failed to assemble into the ribosomes (11). It is conceivable that the defective ribosome biogenesis may have deleterious effects on embryogenesis. Alternatively, mutated S19 could alter the translational efficiency in a mature ribosome, leading to abnormal phenotypes. Although the exact mechanism is still unclear, it is likely that expression of a defective RPS19 protein affects morphogenesis in zebrafish in a way that differs from the absence of RPS19 due to MO knockdown.
Other RP genes potentially involved in hematopoiesis
Many RPs are believed to have additional functions besides their role in ribosomes (32,33). Initially, the loss of such an extraribosomal function of RPS19 was assumed to be the cause of the pure red-cell aplasia in DBA (3). Subsequent studies, however, suggested that the ribosome itself has a role in this disease. If so, involvement of other RP genes in DBA seems plausible. Interestingly, mutations in RPS24 and RPS17 have been identified in some DBA patients who do not have any mutation in RPS19 (12,13). Therefore, some RP genes may have shared functions that are important for hematopoiesis. In this study, the morphants for rpl35, rpl35a and rplp2 genes also displayed a severe reduction in blood cell production. We assume that, like rps19 morphants, the erythropoietic system is impaired in these morphants, suggesting that these RPs have a common role in hematopoiesis. Recently, mutations in RPL35A were reported in DBA patients (34), indicating the usefulness of our data for screening other DBA candidate genes.
Ribosome defects and human diseases
In Drosophila, haploinsufficiency of any RP gene leads to the Minute mutant, which displays thin bristles, poor fertility and an overall developmental delay (35,36). Such phenotypes are believed to be a consequence of decreased translational efficiency due to a reduction of fully functional ribosomes. However, Ts and Bst mutants in mice (37,38) and DBA in humans, which are strongly linked to mutations in RP genes, display tissue-specific phenotypes. It is conceivable that the loss of function of a specific RP may not be critical for all cell types. Accordingly, a lower RPS19 gene dosage may not have the same effect on other cells. In other words, mutations in RPS19 may lead to serious defects in the erythroid lineage but not in other cell lineages. Recently, it was reported that the yeast ribosomes include a subset of paralogous RPs, and specific combinations of RPs are required for translating specific mRNAs (39). Thus, in higher organisms, such functionally specific RPs may impart organ-specific translational preferences to ribosomes, and depletion of an RP may lead to selective effects in a particular organ. Based on this assumption, we hypothesize that, in DBA, the mutations in the RPS19 gene affect the translational efficiency of mRNAs essential for erythroblast differentiation. The RPS19-deficient zebrafish model developed in this study could be a valuable tool to explore this possibility and shed light on the relationship between RPS19 mutations and erythroid cell susceptibility in DBA.
MATERIALS AND METHODS
MOs were obtained from Gene Tools, LLC (Philomath, OR, USA). The MOs were injected into one-cell-stage embryos at a concentration of 0.5 µg/μl. The sequences of the MOs were rps19 MO, 5′-CACTGTTACACCACCTGGCATCTTG, and control MO, 5′-CACTcTTAgACgCACCTGcCATgTTG (bases complementary to the start codon are underlined and mispaired bases are shown in lower case). The sequences of the 19 other RP MOs can be found in our database (http://zebrafish.med.miyazaki-u.ac.jp). These MOs were injected into embryos at the optimal concentrations indicated elsewhere (18).
Whole mount in situ hybridizations
Digoxigenin-labeled antisense riboprobes were transcribed from a linearized plasmid containing gata1 and a PCR-based template containing tie-1 using DIG RNA labeling Mix and T7 RNA polymerase (Roche). The template for the tie-1 probe was generated by PCR using the forward primer 5′-CTGGCCCTCTTTTACATTCG and reverse primer 5′-TAATACGACTCACTATAGGGACTAGGCAGTCTTCCCATGGTTT.
Total RNA was isolated from wild-type and MO-injected embryos using TRIzol reagent (Invitrogen). Semi-quantitative RT–PCR was performed with 0.5 µg total RNA using an OneStep RT–PCR kit (Qiagen). The primer pairs for each gene were as follows: pu.1, 5′-CAGAGCTACAAAGCGTGCAG, and 5′-GCAGAAGGTCAAGCAGGAAC; l-plastin, 5′-GGCATACGGGAGAAAGATGA and 5′-ATGTTGCTGCCCAGTTTAGG; mpo, 5′-AGGGCGTGACCATGCTATAC and 5′-CGGTGTTGTCGCAGATTATG; beta-actin, 5′-GCCCATCTATGAGGGTTACG and 5′-GCAAGATTCCATACCCAGGA.
mRNA synthesis for rescue experiments
Full-length rps19 (GenBank accession no. NM_200750) was amplified by PCR using the forward primer 5′-GCAAGATGCCAGGTGGTGT and the reverse primer 5′-TTATTTTACACTTTCTTGCTTGCAG and cloned into a TA vector (Promega, Madison, WI, USA). Using this TA vector as a template, we generated the cDNA for ‘modified rps19’, which included a silent five-base change around the MO binding site. We used this as a template to synthesize ‘missense rps19’ and ‘nonsense rps19’ mRNAs, corresponding to two mutations found in DBA patients. These cDNAs were then digested by EcoRI and cloned into a pCS2+ vector (provided by Dr Kunio Inoue, Kobe university, Japan) for in vitro transcription. Capped mRNAs were synthesized from 1 µg of the linearized template by SP6 RNA polymerase using a mMessage mMachine kit (Ambion, Austin, TX, USA).
Observation of blood cell circulation and hemoglobin staining
The embryos were grown at 28.5°C and the circulating blood cells in the posterior cardinal vein and the common cardial vein were recorded at 26 and 29 hpf by live video imaging using CCD camera (DP70, Olympus) attached to a stereoscopic microscope (SZX12, Olympus). To examine the density of red blood cells around the cardial vein at 32–48 hpf, we carried out hemoglobin staining using o-dianisidine as previously described (40). We divided the RP morphants into three groups depending on the extent of recovery of the blood cell density and circulation pattern at later stages of development.
This work was supported by Grants-in-Aid (20790734, 20659044, 1806457) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) and Japan Society for the Promotion of Science (JSPS), and funds from The Life Science Foundation of Japan and The Naito Foundation. A.C. is a research fellow of JSPS (P06457).
We thank Dr Kunio Inoue for kindly providing the pCS2+ vector; Dr Noriyoshi Sakai and Dr Minori Shinya for technical advice; Dr Kinta Hatakeyama, Dr Yujiro Asada and Ms Ritsuko Sotomura for useful suggestions; and Dr Maki Yoshihama for useful discussions.
Conflict of Interest statement. None declared.