Whole-exome sequencing and functional studies identify RPS29 as a novel gene mutated in multicase Diamond-Blackfan anemia families

Abstract

Diamond-Blackfan anemia (DBA) is a cancer-prone inherited bone marrow failure syndrome. Approximately half of DBA patients have a germ-line mutation in a ribosomal protein gene. We used whole-exome sequencing to identify disease-causing genes in 2 large DBA families. After filtering, 1 nonsynonymous mutation (p.I31F) in the ribosomal protein S29 (RPS29[AUQ1]) gene was present in all 5 DBA-affected individuals and the obligate carrier, and absent from the unaffected noncarrier parent in 1 DBA family. A second DBA family was found to have a different nonsynonymous mutation (p.I50T) in RPS29. Both mutations are amino acid substitutions in exon 2 predicted to be deleterious and resulted in haploinsufficiency of RPS29 expression compared with wild-type RPS29 expression from an unaffected control. The DBA proband with the p.I31F RPS29 mutation had a pre-ribosomal RNA (rRNA) processing defect compared with the healthy control. We demonstrated that both RPS29 mutations failed to rescue the defective erythropoiesis in the rps29(-/-) mutant zebra fish DBA model. RPS29 is a component of the small 40S ribosomal subunit and essential for rRNA processing and ribosome biogenesis. We uncovered a novel DBA causative gene, RPS29, and showed that germ-line mutations in RPS29 can cause a defective erythropoiesis phenotype using a zebra fish model.

Figure 1

Pedigree of DBA Families. (A) Family NCI-193; (B) Family NCI-38. The diamond indicates the number of healthy siblings not shown.

Figure 2

Diagram of RPS29 conserved domains and genomic structure. (A) Comparison of amino acid conservation of RPS29 homologs; a higher percent identity at a given position is indicated by a deeper blue color, and the 2 sites of mutation are indicated by the boxes. (B) The three-dimensional model of human 40S RPs and RPS29 (UniProt P62273, RS29_HUMAN) was constructed using the UCSF Chimera package. Secondary structural domains are shown as ribbons (α-helix; arrow is β-sheet, and tubes are loop regions), and the amino acid side chains of the RPS29 protein are illustrated. The sites of the amino acid substitutions are shown and highlighted in red. (C) Schematic of RPS29 secondary structure (black regions are coils, the green region is a helix, and blue regions are strands) with the sites of the amino acid substitutions shown.

Figure 3

Functional assays for the RPS29 mutations. (A) Quantitative RT-PCR for RPS29 expression was performed on peripheral blood–derived lymphoblasts and primary fibroblast RNA samples from the proband from family NCI-193 (p.I31F) and family NCI-38 (p.I50T) and an unaffected control individual WT RPS29. The figure shows the combined data from 3 plates for lymphoblasts, and data for 1 plate for the fibroblasts, normalized using 2 endogenous controls (ACTB and GAPDH), for the short and long isoforms of RPS29. (B) Northern blot analysis of pre-rRNA processing was performed with a hybridization probe to 18SE/ITS1 (probe γ) using activated lymphocytes recovered from the peripheral blood from the proband (III-3) in family NCI-193 (p.I31F) and an unaffected control individual with WT RPS29.

Figure 4

Zebra fish as a DBA model of the RPS29 mutations. (A) The Hb phenotype of zebra fish embryos with WT rps29^+/+, rps29^−/− uninjected embryos, rps29^−/− embryos injected with WT human RPS29 RNA, and rps29^−/− embryos injected with the p.I31F mutant human RPS29 RNA from the proband (III-3) in family NCI-193. The numbers of zebra fish embryos displaying this phenotype are shown (ie, 9 of 25 injected and 21 of 21 injected). (B) The percent of rps29^−/− zebra fish embryos with high Hb are shown for the uninjected embryos, WT human RPS29 RNA–injected embryos, and the p.I31F mutant human RPS29 RNA–injected embryos. (C) The Hb phenotype of embryos injected with the p.I50T mutant human RPS29 RNA. (D) The percent of embryos with high Hb. *Student t test P < .05; **Student t test P < .01.