Diamond-Blackfan anemia: genotype-phenotype correlations in Italian patients with RPL5 and RPL11 mutations

Abstract

Background Diamond-Blackfan anemia is a rare, pure red blood cell aplasia of childhood due to an intrinsic defect in erythropoietic progenitors. About 40% of patients display various malformations. Anemia is corrected by steroid treatment in more than 50% of cases; non-responders need chronic transfusions or stem cell transplantation. Defects in the RPS19 gene, encoding the ribosomal protein S19, are the main known cause of Diamond-Blackfan anemia and account for more than 25% of cases. Mutations in RPS24, RPS17, and RPL35A described in a minority of patients show that Diamond-Blackfan anemia is a disorder of ribosome biogenesis. Two new genes (RPL5, RPL11), encoding for ribosomal proteins of the large subunit, have been reported to be involved in a considerable percentage of patients.Design and Methods In this genotype-phenotype analysis we screened the coding sequence and intron-exon boundaries of RPS14, RPS16, RPS24, RPL5, RPL11, and RPL35A in 92 Italian patients with Diamond-Blackfan anemia who were negative for RPS19 mutations.Results About 20% of the patients screened had mutations in RPL5 or RPL11, and only 1.6% in RPS24. All but three mutations that we report here are new mutations. No mutations were found in RPS14, RPS16, or RPL35A. Remarkably, we observed a higher percentage of somatic malformations in patients with RPL5 and RPL11 mutations. A close association was evident between RPL5 mutations and craniofacial malformations, and between hand malformations and RPL11 mutations.Conclusions Mutations in four ribosomal proteins account for around 50% of all cases of Diamond-Blackfan anemia in Italian patients. Genotype-phenotype data suggest that mutation screening should begin with RPL5 and RPL11 in patients with Diamond-Blackfan anemia with malformations.

Introduction

Diamond-Blackfan anemia (DBA, MIM#105650) is a rare inherited congenital bone-marrow-failure syndrome characterized by normochromic macrocytic anemia typically presenting in infancy or early childhood. The bone marrow is normocellular, but erythroid precursors are absent or their numbers markedly decreased, because their progenitors are unable to differentiate and are prone to apoptosis. Laboratory findings such as increased mean corpuscular volume, high erythrocyte adenosine deaminase activity (eADA), and elevated hemoglobin F after 6 months of age are observed in most patients; an increase in eADA may be the only manifestation.1 The congenital anomalies, mainly involving the head, upper limbs, heart and urogenital system, found in more than one-third of patients reflect the fact that DBA is a broad disorder of development.2 The range of severity of such malformations is wide, even within the same family.

The genetic basis of DBA is heterogeneous. Approximately 40% of patients have mutations in one of the genes for ribosomal proteins (RP): RPS7, RPS17, RPS19, RPS24, RPL5, RPL11, or RPL35A. These genes encode for RP of either the small or the large ribosomal subunit.3–7 The identification of the role of another RP gene, RPS14, in the pathogenesis of an acquired myelodysplastic disease, the 5q- syndrome, has recently attracted great interest, since it extends the spectrum of ribosomal diseases.8

A clear genotype-phenotype correlation is not apparent in patients with RPS19 mutations, and identical mutations have been found in patients with a wide range of clinical presentations, even within the same family. No information of genotype-phenotype correlations is so far available for patients with RPS24, RPS17, or RPL35A mutations, because the number of subjects studied are too small.3–5 On the other hand a correlation between RPL11 or RPL5 mutations and hand malformations and cleft lip and/or palate, respectively, has been reported.6,7

Here we report the results of screening for six RP genes (RPS14, RPS16, RPS24, RPL5, RPL11, and RPL35A) in 92 unrelated Italian patients who were negative for RPS19 mutations, and a genotype-phenotype analysis for the RPL5 and RPL11 genes.

Design and Methods

Patients

One hundred twenty-eight unrelated DBA families were studied. Fourteen had more than one clinically affected individual. The diagnosis of DBA was always based on normochromic, often macrocytic anemia, reticulocytopenia, erythroid bone marrow aplasia or hypoplasia, and, in some patients, congenital malformations and elevated eADA. We excluded short stature because it was difficult to evaluate in the context of severe anemia, iron overload and chronic corticosteroid use.

Informed consent was obtained from all patients and/or their family members participating in the study. RPS19 mutations were found in 36/128 (28%) unrelated DBA patients using both sequencing and multiplex ligation-dependent probe analysis (MLPA): these data have already been reported.9,10 Specifically, 33 RPS19 mutations were identified by sequencing whereas three heterozygous RPS19 deletions missed by sequencing were found using the MLPA technique.10

Molecular analysis of RP genes

Genomic DNA from 92 unrelated Italian DBA probands negative for RPS19 mutations was isolated from peripheral blood leukocytes using a commercial kit (Gentra Systems, Inc., Minneapolis, MN, USA). We analyzed RPS24, RPL5, RPL11 and RPL35A because mutations in these genes have been reported in the literature.3,4,6 We screened RPS16 because it is involved in binding of initiation factor eIF-2 to the 40S subunit.11 The RPS14 gene was screened because it has an important role in erythroid proliferation and maturation.8 This gene, together with others, is deleted in the 5q- syndrome, an acquired myelodysplastic syndrome which is very different from DBA, but considered a ribosomal disease.8

We analyzed these six RP genes by direct sequencing of the coding exons and intron-exon boundaries. The primer sequences are available on request. Polymerase chain reaction (PCR) products were purified with the QIAquick purification kit (QIAGEN GmbH, D-40724 Hilden, Germany), and sequenced on both strands with an ABI PRISM BigDye Terminator kit (Applied Biosystems, Foster City, CA, USA) on an Applied Biosystems 3100 DNA Sequencer (Applied Biosystems). When sequence changes were found, independent PCR products were sequenced to confirm the mutations.

Subsequently, we sequenced DNA samples from available family members to determine whether the mutation co-segregated with the DBA phenotype within the pedigree. To determine whether these sequence changes were polymorphic variations, we sequenced DNA samples from 100 Italian control individuals, and verified that none was reported in the Single Nucleotide Polymorphism database (dbSNP at www.ncbi.nlm.nih.gov/SNP) or in the Ensembl database (www.ensembl.org). To verify whether the RPS24 missense mutation (p.Asn124Ser) was a polymorphism we performed ScrFI enzymatic digestion on samples from 150 Italian control individuals.

The nomenclature used to describe the sequences is in accordance with the Human Genome Variation Society recommendations (http://www.hgvs.org).

To ascertain whether the two RPL5 mutations detected in patient 2 (Table 1) were on the same allele a test based on NlaIII restriction enzyme digestion was set up. The splice site mutations detected in patients 10 and 12 (Table 1) were analyzed with a Splice Site Prediction website (BDGP, Berkeley Drosophila Genome Project, http://www.fruitfly.org).

Cell culture and transfection

Human embryonic kidney (HEK) 293T cells (ATCC #CRL-11268) were cultured in Dulbecco’s modified essential medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin and 100 μg/mL streptomycin at 37°C with 5% CO₂. For transfection, RPS24 cDNA was reverse transcribed and amplified from total RNA using SuperScript® III (Invitrogen, San Diego, CA, USA). Both variants 1 and 2 (NM_033022 e NM_001026) were obtained and inserted in a pcDNA3.1+ vector (Invitrogen) that contained the neomycin resistance gene and was modified to express a Flag tag at the C-terminus. Site-directed mutagenesis for mutant c.371A>G (p.Asn124Ser) was carried out using PfuTurbo® DNA Polymerase (Stratagene, La Jolla, CA, USA) and the following primers: forward 5′-tgcaaaggccagtgttggtgctg-3′, reverse 5′-cagcaccaacactggcctttgca-3′. A PCR-based strategy was applied to obtain mutant 64_66delCAA, using Platinum®Pfx DNA Polymerase (Invitrogen) and the primers, forward 5′-Phos-atggt-cattgatgtccttcaccc-3′, reverse 5′-tttcctctgaagtagtcggttggt-3′. All constructs were verified by DNA sequencing. About 2 μg of each plasmid and 20 μL of lipofectamine transfection reagent (Invitrogen, Lipofectamine plus 2000) were used for transfection of HEK293 cells plated at 90% confluence. After 24 h cells were analyzed.

Table 1.RPL5 mutations.

RPS24 protein analysis

To prepare total extracts, cells were washed twice with phosphate-buffered saline (150 mM NaCl, 2.7 mM KCl, 8 mM Na₂PO₄ and 1.4 mM KH₂PO₄) and treated with lysis buffer (150 mM NaCl, 50 mM Tris–HCl (pH 7.5), 1% NP40, 0.5% sodium deoxycholate, 0.1% sodium dodecylsulfate [SDS], aprotinin 1 mg/mL, leupeptins 1 mg/mL, pepstatin A 1 mg/mL, phenylmethylsulfonylfluoride 100 mg/mL). After 1 min of incubation on ice, the extract was centrifuged for 10 min at maximum speed in a microcentrifuge at 4°C. For fractionation of the extract, cells were lysed in 300 μL of 10 mM Tris HCl (pH 7.5), 10 mM NaCl, 3 mM MgCl₂, 0.05% NP40 with protease inhibitors and centrifuged for 5 min at 1000 g. To prepare the nuclear fraction, the pellet was resuspended in 1.2 mL of 10 mM Tris HCl (pH 7.9), 10 mM NaCl, 5 mM MgCl₂, 0.3 M sucrose and layered onto a 2 mL cushion of 0.6 M sucrose. After 10 min centrifugation at 1,000g at 4°C, the pellet was dissolved in SDS–polyacrylamide gel electrophoresis (PAGE) loading buffer as a nuclear fraction. The supernatant of the first centrifugation at 1000 g was either used as cytoplasmic fraction or further purified. To prepare the ribosomal fraction the cytoplasmic extract was layered onto 1 mL of 15% sucrose, 30 mM Tris–HCl (pH 7.5), 100 mM NaCl, 10 mM MgCl₂ and centrifuged in a Beckman type 70.1 rotor for 90 min at 100,000g. The pellet (ribosomal fraction) was resuspended directly in SDS–PAGE loading buffer. The supernatant (free cytoplasmic proteins) was precipitated with 10% trichloroacetate and the pellet, washed with acetone, was resuspended in SDS–PAGE loading buffer. For western analysis, proteins were separated on a 12% SDS polyacrylamide gel, transferred to a nitrocellulose Protran membrane (Schleicher and Schuell), and incubated with the following primary antibodies and antisera: mouse monoclonal anti-RPS19,12 mouse anti-Flag (Sigma, F3165), rabbit anti-NPT II (Upstate, 06-747), and mouse anti-GAPDH (Millipore MAB374). Primary antibodies were revealed using horseradish peroxidase-conjugated goat anti-rabbit antibody (Jackson Immunoresearch) and a chemiluminescence detection system (Ablot Plus, Euroclone). Quantitation analyses were performed using the LAS3000 Image System (Fuji) and ImageQuant software (GE Healthcare).

Genotype-phenotype correlations and statistical analysis

Genotype-phenotype correlations were evaluated in the whole group of patients. Differences between two independent samples were checked with Student’s t-test or the Mann-Whitney test as appropriate, whereas the Kruskal-Wallis test was used to assess the differences between more than two groups. Associations between categorical variables were assessed with Fisher’s exact test or with odds ratio and 95% confidence interval (95% CI). For categorical variables with more than two categories simple logistic regression was used to calculate the odds ratio and 95% CI.

All tests were two-sided and P values less than 0.05 were considered statistically significant. Data were analyzed with the SPSS 16 software (SPSS Inc., Chicago, IL, USA).

Results

RPL5

We identified RPL5 sequence changes in 12 out of the 92 probands (Table 1). These included seven deletions or insertions of one to five nucleotides causing a frameshift, three donor splice-site mutations, one missense and one nonsense mutation. All mutations but one were new mutations not yet described in the literature and were found in one patient only. The four-nucleotide deletion in exon 3 (c.169_172delAACA) detected in two unrelated patients has already been reported.6,7 De novo sequence changes were identified in six probands; one patient (n. 10, Table 1) inherited the RPL5 mutation (c.3+3G>C) from his healthy mother, who had normal hemoglobin level, mean corpuscular volume and eADA. Parental DNA was not available for the other five patients.

Table 2.RPL11 mutations.

A five-nucleotide deletion in exon 3 (c.134_138delACACA) was found in a female patient (n. 3, Table 1), who also carried a RPS24 missense mutation in exon 4 causing a substitution of an Asparagine with a Serine at codon 124 (c.371A>G; p.Asn124Ser). This patient is steroid-dependent, has a cleft palate, short stature and increased eADA. The RPL5 mutation was de novo, while the RPS24 mutation was found in her healthy father and unaffected sister (Figure 1; Table 1, n. 3).

Patient n. 2 (Table 1) had two mutations on the same allele. The first was a nonsense mutation causing a premature stop codon (p.Tyr226X). The second was a missense mutation involving the adjacent codon 227 (p.Ile227Arg), and caused the substitution of an Isoleucine with an Arginine. Parental DNA was not available for molecular analysis.

The splice site mutations detected in patients n. 10 (c.3+3G>C; Table 1) and n. 12 (c.324+5G>T; Table 1) modified the splice site strength score from 0.97 and 0.98 to 0.32 and 0.23, respectively (BDGP, http://www.fruitfly.org). These data suggested splice site suppression in both mutants. We could not perform expression analysis because fresh blood samples from these patients were not available.

RPL11

We identified sequence changes in 12 patients (Table 2). Most mutations (11/12, 92%) were 1-47 nucleotide deletions. An acceptor splice-site mutation was also observed. Nine mutations are described here for the first time; two, including one present in two unrelated probands, had been reported by Gazda et al.6 Parental DNA was available for nine probands; seven had de novo mutations. Two probands (n. 2 and n. 12, Table 2) inherited the mutation from their apparently healthy mother and affected father, respectively. This father reported mild anemia during infancy associated with short stature, increased mean corpuscular volume and high eADA; he has never needed any treatment, whereas his daughter is transfusion-dependent, and does not respond to steroids.

RPS24

We found two new heterozygous changes in RPS24 (2/92). A deletion of three nucleotides in exon 2 (c.64_66delCAA), resulting in the loss of the highly conserved glutamine 22, was identified in a patient without somatic malformations and in clinical remission at last follow-up. This mutation was not found in other family members, nor has it been reported as a polymorphism (www.ncbi.nlm.nih.gov/SNP; www.ensembl.org).

A missense mutation (p.Asn124Ser) in exon 4 was found in a patient who also carried a five-nucleotide deletion (c.134_138delACACA) in RPL5 as described above (Figure 1; Table 1 n. 3). This missense mutation was carried by two healthy family members (Figure 1; Table 1, n. 3), but has never been reported as a polymorphism nor was it found in 300 Italian control chromosomes.

Neither RPS24 mutation predicted a dramatic alteration of the gene product, differently from RPL5 and RPL11 mutations. Thus, we decided to study the properties of RPS24 proteins encoded by the mutated genes, using an approach similar to that which we previously used to study RPS19 missense mutations.12 We prepared cDNA constructs tagged with the Flag epitope at the C terminus and inserted them into mammalian expression vectors (pcDNA3.1). RPS24 pre-mRNA is alternatively spliced into three different mRNA isoforms that encode proteins with a slightly different C terminus (one or three additional amino acids).3 We prepared cDNA encoding the two major RPS24 variants (1 and 2) but, since in a preliminary analysis they showed identical behaviors (data not shown), we focused our attention on one protein variant (variant 1). DNA constructs encoding for: (i) wild type RPS24 (WT), (ii) RPS24 with codon 22 deletion (M1), and (iii) RPS24 with the p.Asn124Ser mutation (M2) were used in transient transfection experiments into HEK293 cells. After transfection, cell extracts were analyzed by western blotting to measure protein levels or further purified to investigate the subcellular localization and the capacity to assemble into the ribosome. The levels of WT, M1 and M2 RPS24 were normalized for the amount of neomycin phosphotransferase II expressed by the pcDNA3.1 vector. Assuming that the different RPS24 cDNA have the same transcriptional and translational activity, the results can be considered proportional to the stability of the different proteins. As shown in Figure 2A the levels of both M1 and M2 are clearly lower than that of WT RPS24. This indicates that both mutations affect protein stability. Consistent with the instability of the mutated RPS24, a possible degradation product appeared sometimes on the western blot as a faster-migrating band of variable intensity. Next, we separated nuclear and cytoplasmic fractions from the extract of the transfected cells. Western blot analysis, illustrated in Figure 2B, showed that mutated RPS24 (M1 and M2) accumulated into the nucleus more evidently than did the WT RPS24. Finally, to verify the capacity of the mutated RPS24 to be incorporated into the ribosome, we further fractionated cytoplasmic extracts through ultracentrifugation on a sucrose cushion. After 2 h of centrifugation all ribosomes and ribosomal subunits were found in the pellet whereas free cytoplasmic proteins could be recovered from the supernatant. As an additional control for this experiment, we transfected a plasmid expressing an RPS19 construct (RPS19flag) with the Flag epitope at the N terminus. The RPS19flag fusion protein was previously shown to assemble poorly into the ribosome,12 possibly as a consequence of the position of the Flag epitope (N terminus).

Figure 1.Pedigree of a DBA patient carrying mutations in both the RPL5 and RPS24 genes.

Western blot analysis, presented in Figure 2C, confirmed that RPS19flag was mostly in the free cytoplasmic fraction. In contrast, both the WT and the mutated RPS24 (M1 and M2) appeared to be mainly associated with the ribosome. This result indicates that, although the mutated RPS24 are less stable and accumulate into the nucleus, a small fraction of them is incorporated into the ribosome and exported into the cytoplasm. The conclusion of our analyses is that the two mutations here analyzed alter some properties of RPS24 in a similar way. However, a fraction of the mutated RPS24 was able to associate with the ribosome.

RPS14, RPS16 and RPL35A

No mutations were found in RPS14, RPS16, or RPL35A.

The clinical features of patients with RPL5 and RPL11 mutations are summarized in Tables 1 and 2. The results of the genotype-phenotype statistical analysis performed by logistic regression are presented in Figure 3.

Most of the patients with RPL5 (83%) and RPL11 (73%) mutations had physical malformations. Patients with RPS19 mutations and patients without mutations in any of these three genes showed lower percentages of malformations (43% and 29%). Specifically, the risk of malformation was 12-fold higher in RPL5-mutated patients than in patients with no mutations in RPL5, RPL11 or RPS19 and 7-fold higher in RPL5-mutated patients than in RPS19-mutated patients. Similarly, RPL11-mutated patients had a 6-fold higher and 3.5-fold higher risk of somatic anomalies compared to RPL5, RPL11 and RPS19 non-mutated and RPS19-mutated patients. Moreover, patients with either RPL5 or RPL11 mutations more frequently had multiple malformations than did RPL5, RPL11 and RPS19 non-mutated and RPS19-mutated patients.

We also found that craniofacial abnormalities were closely associated with RPL5 mutations: RPL5-mutated patients had a 37-fold and an 8-fold higher risk of craniofacial malformations than RPL5, RPL11 and RPS19 non-mutated and RPS19-mutated patients, respectively. RPL11-mutated probands had a higher risk of craniofacial malformation than RPL5, RPL11 and RPS19 non-mutated patients only. Interestingly, cleft lip and/or palate was observed in 8/111 clinically evaluable patients. None of these eight had a mutation in RPS19, whereas all had mutations in RPL5 or RPL11.

A significant association between hand malformations and RPL11 mutations was also observed: RPL11-mutated patients had a 13-fold and a 7-fold higher risk of hand malformation compared to RPL5, RPL11 and RPS19 non-mutated and RPS19-mutated patients, respectively. A similar, though weaker, association was found in RPL5-mutated patients compared to RPL5, RPL11 and RPS19 non-mutated patients; no difference was observed between RPS19 and RPL5-mutated patients.

The risk of cardiac malformations was 5-fold and 6-fold higher in RPL5 and RPL11-mutated patients compared to RPL5, RPL11 and RPS19 non-mutated patients only. We did not find a correlation between mutational status and response to first steroid treatment or status at last follow-up.

Figure 2.Transient transfection of RPS24 constructs. (A) Total extracts from HEK293 cells transfected with plasmids coding for RPS24 WT (WT) and mutated forms of the protein (M1, M2) were separated by SDS-PAGE and transferred on nitrocellulose membrane. The blots were decorated with anti-Flag (Flag) and anti-neomycin phosphotransferase II (NPT) antibodies. Quantitation of the signals is reported in the lower part as the ratio of Flag/NPT; (B) Total extracts from transfected cells (as in A) were separated into nuclear (N) and cytoplasmic (C) fractions. The loading ratio between cytoplasmic and nuclear extracts was 1:20 (number of cells). Western blotting was performed as in A with anti-Flag (Flag) and anti-RPS19 (RPS19). (C) Cytoplasmic extracts from HEK293 cells transfected as in A were fractionated by ultracentrifugation on a sucrose cushion as indicated in the Design and Methods section. Pellets, which include ribosome and ribosomal subunits (P) and trichloroacetate-precipitated supernatants, which include free cytoplasmic proteins (S) were analyzed by western blot with anti-Flag (Flag) and anti-glyceraldehyde phosphate dehydrogenase (GAPDH) antibodies.

Discussion

For about 10 years, RPS19 seemed to be the only gene involved in the pathogenesis of DBA and mutations in this gene accounted for 25% of cases.9 In the last 3 years, however, heterozygous mutations in several genes encoding ribosomal proteins of either the small or the large ribosomal subunit have been reported.3–7 Specifically, Gazda et al. found mutations in RPL5 and RPL11 in about 7% and 5%, respectively, of DBA patients,6 while Cmejla et al. showed a higher frequency of RPL5 (21%) and RPL11 (7%) mutations among Czech patients.7

We report here the results of our screening of RPS24, RPL5 and RPL11 in 92 Italian patients who were negative for RPS19 mutations. No mutations were found in RPS14, RPS16, and RPL35A.

Twenty-eight percent of our patients (36/128) showed a mutation in RPS19,9,10 9.3% (12/128) in RPL5, 9.3% (12/128) in RPL11, and only 1.6% (2/128) in RPS24. This frequency of RPL11 mutations was higher than that found in other screening studies. As far as concern RPL5 mutations, the frequency in our study was similar to that found by Gazda et al. but lower than in Czech patients.6,7 These differences may be due to the populations studied. All but three mutations found in these genes had never been previously described. Interestingly, we found a double heterozygote for two RP genes. This patient, a steroid-dependent female with cleft palate, had a de novo five-nucleotide deletion in RPL5 exon 3 (c.134_138delACACA) and a substitution in RPS24 (p.Asn124Ser). The latter was inherited from her healthy father, and also carried by a healthy sister (Figure 1). Several points support the hypothesis that the RPL5 mutation is pathogenic: it is a de novo frameshift mutation and the malformation phenotype of this patient is typical of that of patients with a RPL5 mutation. There is, however, controversy on the role of the missense RPS24 p.Asn124Ser: this mutation is carried by healthy relatives and Asparagine 124 is a non-conserved amino acid. However, a silent phenotype has already been described in DBA patients with RPS24 mutations.3 p.Asn124Ser was not found in 300 Italian normal chromosomes and has never been reported as a single nucleotide polymorphism.

To ascertain whether p.Asn124Ser is a mutation or a silent polymorphism we checked whether this mutant was able to reach the nucleolus and be included in active ribosomes. Our studies show that though the mutant is rather unstable as compared with the wild type form, it is able to reach the nucleolus and a small fraction is even able to associate with the ribosome. Altogether, our studies support the following model: the bulk of the DBA phenotype is most probably due to the RPL5 frameshift mutation, whereas the RPS24 missense mutation is a phenotype modifier.

Interestingly, similar functional data were found for the in frame RPS24 mutation: however, in this case, the fact that this is a de novo mutation is a strong indication that this deletion has a role in the pathogenesis of the disease.

Overall, our data suggest that all DBA patients, even those with a recognized mutation in RP genes, should be screened for other DBA genes to ascertain whether a second mutation is present. A possible digenic effect may explain the variable expressivity and incomplete penetrance in members of the same family.

Figure 3.Malformation status of patients with RPL5 and RPL11 mutations. Associations between malformation status and RP gene mutations were assessed with odds ratio (OR) and 95% CI calculated from logistic regression; OR are drawn on a logarithmic scale. RPL5 and RPL11-mutated patients were compared to both RPS19-mutated patients and non-mutated patients. The Wald test was used to test the statistical significance of each association, the P value is indicated as (*) P<0.05; (§) P<0.01.

Incomplete penetrance was shown for RPL5 and RPL11 mutations in two families of our cohort (n. 10, Table 1; n. 2, Table 2), as previously described for RPS19.13 All RPS19 mutations described have been found in heterozygosis, suggesting haploinsufficiency due to a loss-of-function effect.9 Similarly, most RPL5 and RPL11 mutations cause a premature termination, suggesting that haploinsufficiency is the cause of the disease.

Gazda et al. reported that RPL5 and RPL11 mutations are more frequently associated with physical malformations than are RPS19 mutations. RPL5 mutations seem to cause a more severe malformation status, including craniofacial, thumb, and heart anomalies, compared to mutations in RPS19.6 Remarkably, Gazda et al. reported a close association between RPL5 and cleft lip and/or palate. Mutations in RPL11 were predominantly associated with thumb abnormalities.6 Similarly, Cmejla et al. observed that all ten Czech DBA patients with either an RPL5 or an RPL11 mutation had one or more physical malformations; specifically, hand anomalies were always present and most patients were born small for gestational age.7 Cleft lip and/or palate was not observed in RPL5-mutated and RPL11-mutated patients.

We observed a high percentage of multiple somatic malformations in patients with RPL5 and RPL11 mutations. Close associations between RPL5 mutations and craniofacial malformations and between hand malformations and RPL11 mutations were also evident. However, unlike Gazda et al., we observed cleft lip and/or palate in both RPL5- and RPL11-mutated DBA patients, and can thus rule out a unique association between RPL5 mutations and this type of malformation.

In conclusion, we report a high frequency of RPL5 (9.3%) and RPL11 (9.3%) mutations in our DBA cohort. Mutations in four ribosomal proteins account for around 50% of all cases of DBA in Italian patients. Genotype-phenotype data suggest that mutation screening should begin with RPL5 and RPL11 in DBA patients with a malformation, and specifically in those who have craniofacial aberrations, such as cleft palate, or hand abnormalities.

Footnotes

• Funding: this work was supported by grants from Banca del Piemonte to UR, PRIN 2006 Ministero Italiano della Ricerca Scientifica e Tecnologica to UR and ID, Regione Piemonte to UR, Gruppo di Sostegno DBA Italia to UR, DBA Foundation to ID, and Telethon Italia Grant GGP07242 to ID.
• We thank the Daniella Maria Arturi Foundation for supporting communication among DBA researchers.
• Authorship and Disclosures PQ performed genotype-phenotype correlation analyses and drafted the manuscript with UR and ID, who were involved in the conception of the study. UR is the author taking primary responsibility for the paper. EG and AC performed sequencing analyses. AB critically revised the paper and was responsible for important intellectual content. RC was responsible for the statistical analysis. CD, DL, AM and LV contributed clinical data. AA performed cloning. LB and FL performed RPS24 functional experiments. All authors contributed to the revision of the manuscript.
• The authors declare no potential conflicts of interest.

• Received May 27, 2009.
• Revision received July 26, 2009.
• Accepted August 5, 2009.