Functional Consequences Of RPS29 Germline Mutations In Diamond-Blackfan Anemia

Diamond-Blackfan anemia (DBA) is an inherited bone marrow failure syndrome (IBMFS) characterized by red blood cell aplasia, variable physical anomalies, and increased risks of cancer. Approximately half of DBA cases have a germline mutation or deletion in a ribosomal protein gene (i.e., RPS19, RPS24, RPS17, RPL35A, RPL5, RPL11, RPS7, RPS10, or RPS26), or in GATA1, an X-linked hematopoietic transcription factor. The inheritance of mutated ribosomal protein genes is autosomal dominant (AD), but de novo germline mutations may also occur. We previously reported a novel nonsynonymous ribosomal protein S29 (RPS29) mutation, I31F, which segregated with DBA in a large multi-case family (NCI-193). We now present functional data that link this RPS29 mutation to the DBA phenotype and report an additional large multi-case DBA family (NCI-38) with another novel RPS29 mutation.

Individuals in this study are participants in the IRB-approved longitudinal cohort IBMFS study at the NCI. We have identified a second large family with DBA, NCI-38, in which the male proband had steroid-responsive anemia as a child, elevated erythrocyte adenosine deaminase (eADA = 1.26 IU/g Hb, normal <0.96), and has been in remission for approximately 20 years. Two unaffected male half-siblings have no symptoms of DBA, and 1 female half-sibling has only a borderline elevated eADA (1.00 IU/g Hb). The proband’s mother is an asymptomatic obligate carrier because her full brother and two half-siblings have DBA or an unspecified bone marrow failure; her eADA is normal (0.8 IU/g Hb). We performed whole-exome sequencing using Nimblegen v2 capture arrays and paired-end sequencing on the Illumina HiSeq^TM. After applying quality control filters, removing variants present in publicly available databases (1000Genomes, ESP, Kaviar, and dbSNP), and applying an AD inheritance model, we identified an I50T mutation in RPS29 that tracked with AD inheritance in this family (NCI-38). The sibling with elevated eADA and the mother also carried this RPS29 mutation while the two asymptomatic healthy siblings were mutation negative. I50T, like I31F, occurs in a very highly evolutionarily conserved region of the RPS29 protein and in silico prediction programs (i.e., SIFT, Polyphen 2, Condel, MutationTaster) predict this variant to be deleterious.

The original DBA family, NCI-193, has AD DBA due to the heterozygous I31F mutation in RPS29. We used quantitative rtPCR to determine RPS29 expression in the proband with the I31F mutation compared with a healthy individual with wildtype RPS29. The proband demonstrated haploinsufficiency of RPS29 mRNA, with levels consistent with approximately 50% expression compared with wildtype RPS29 values. Zebrafish are ideal for modeling ribosomal protein knockdown because hematopoietic regulation is conserved with mammals, they are amenable to high throughput in vivo genetics, and ribosomal proteins can be knocked down in the embryo with morpholino technology. The rps29-/- mutant zebrafish displays many aspects of the DBA phenotype including significant defects in red blood cell development, shown by a decrease in hemoglobin levels. We used this as a model to determine if the RPS29 mutation identified in our DBA cases could rescue the hemoglobin defect. The rps29-/- zebrafish embryos were injected with 50pg of either wildtype or the mutated (I31F) human RPS29 RNA. The wildtype human RPS29 RNA rescued the hemoglobin phenotype (increased hemoglobin levels) of the rps29-/- zebrafish embryos, whereas the mutated human RPS29 RNA could not (no increase in hemoglobin levels; P <0.01).

In summary, germline mutations in RPS29 can cause AD DBA. RPS29 is an essential protein in the 40s small subunit of the ribosome, important for ribosomal RNA processing and ribosome biogenesis. Using functional assays, we show for the first time that the mutant RPS29 in our DBA cases results in haploinsufficiency of RPS29. The patient-associated I31F RPS29 mutation failed to rescue the defective hematopoiesis in the rps29-/- mutant zebrafish DBA model and provides further evidence that RPS29 is a DBA-associated gene.