Identification of RPS14 as a 5q- syndrome gene by RNA interference screen

Abstract

Somatic chromosomal deletions in cancer are thought to indicate the location of tumour suppressor genes, by which a complete loss of gene function occurs through biallelic deletion, point mutation or epigenetic silencing, thus fulfilling Knudson's two-hit hypothesis. In many recurrent deletions, however, such biallelic inactivation has not been found. One prominent example is the 5q- syndrome, a subtype of myelodysplastic syndrome characterized by a defect in erythroid differentiation. Here we describe an RNA-mediated interference (RNAi)-based approach to discovery of the 5q- disease gene. We found that partial loss of function of the ribosomal subunit protein RPS14 phenocopies the disease in normal haematopoietic progenitor cells, and also that forced expression of RPS14 rescues the disease phenotype in patient-derived bone marrow cells. In addition, we identified a block in the processing of pre-ribosomal RNA in RPS14-deficient cells that is functionally equivalent to the defect in Diamond-Blackfan anaemia, linking the molecular pathophysiology of the 5q- syndrome to a congenital syndrome causing bone marrow failure. These results indicate that the 5q- syndrome is caused by a defect in ribosomal protein function and suggest that RNAi screening is an effective strategy for identifying causal haploinsufficiency disease genes.

Figure 1

Screen of common deleted region for the 5q- syndrome. Each gene was targeted by multiple lentivirally expressed shRNAs in CD34+ cells from umbilical cord blood, and the ratio of megakaryocytic to erythroid differentiation was determined by flow cytometry using antibodies against CD41 and GlyA respectively. Controls are shown in the left panel, including an shRNA targeting luciferase, which is not expressed in the primary cells, and multiple shRNAs targeting GATA-1, an erythroid-specific transcription factor. All of the genes in the CDR for the 5q- syndrome are shown in the right panel. The megakaryocytic/erythroid ratio is shown as a z-score using the mean and standard deviation of the control (Luc) replicates. For the control shRNA targeting the luciferase gene, circles represent 30 individual replicates. For all other genes, circles represent the median of three replicates for each individual shRNA. The mean of all shRNAs targeting a given gene is shown with a grey bar.

Figure 2

Multiple shRNAs targeting RPS14 recapitulate the 5q- syndrome in vitro. Western blots demonstrate that five different shRNAs effectively decrease levels of RPS14, as shown by Western blot (panel A). Compared to a control shRNA targeting the luciferase gene (Luc), each of the 5 RPS14 shRNAs block erythroid relative to megakaryocytic differentiation in adult bone marrow CD34+ cells. The ratios of cells from the erythroid and megakaryocytic lineages, indicated on the y-axis, were assessed by flow cytometry using antibodies against GlyA and CD41 respectively (panel B). In addition, RPS14 shRNAs decrease erythroid relative to myeloid differentiation, assessed using antibodies against GlyA and CD11b (panel C); block terminal erythroid differentiation, assessed using antibodies against GlyA and CD71 (panel D); and increase apoptosis, assessed by annexin V expression (panel E). In panels B through E, the effect of RPS14 shRNAs, compared to the luciferase shRNA, was statistically significant (p < .05 by Student's two tailed t-test, n=3, mean and S.E.M. shown). Multiple shRNAs targeting RPS14 also alter the transcriptional programs of lineage specific differentiation. The top 100 marker genes that are differentially expressed between cells expressing control versus RPS14 shRNAs, ranked by signal to noise ratio, are shown in panel F. RPS14 is at the top of the list of downregulated genes and is expressed at approximately 40% of the normal level. RPS14 shRNAs significantly decrease expression of an erythroid gene expression signature and increase expression of neutrophil and platelet signatures (panels G – I), as assessed by GSEA. Genes are ranked by signal to noise ratio according to their differential expression between cells expressing RPS14 versus control shRNAs. Genes in the lineage-specific gene sets are marked with vertical bars, and the enrichment score is shown in green.

Figure 3

RPS14 is required for 18S pre-rRNA processing and 40S ribosomal subunit formation. A simplified schematic of pre-rRNA processing is illustrated in panel A. A defect in the 5’ processing of 18S pre-rRNA is evident from Northern blots using RNA from TF-1 cells expressing control or RPS14 shRNAs with an accumulation of 30S rRNA and a deficiency of 21S and 18SE pre-rRNAs and mature 18S rRNA (panel B). The Northern blot probes are shown in red. Polysome profiles from TF-1 cells demonstrate that decreased expression of RPS14 results in a 40S subunit deficiency (panel C). The 30S/18SE pre-rRNA ratio is also increased in RNA from bone marrow mononuclear cells from MDS patients with the 5q deletion (n=4) compared to MDS patients without 5q deletions (n=5), as measured by quantification of Northern blots (p=.06, panel D).

Figure 4

RPS14 overexpression rescues erythroid differentiation in samples from patients with 5q deletions. CD34+ cells from bone marrow aspirates of patients with the 5q- syndrome (shown in red) and MDS patients without 5q deletions (shown in blue) were infected with a lentivirus expressing the RPS14 cDNA or an empty vector. In patients with 5q deletions, RPS14 overexpression increased erythroid relative to megakaryocytic differentiation (panels A and C) and erythroid relative to myeloid differentiation (panels B and D) shown normalized to the empty vector control. The mean and standard deviation of three independent experiments are shown. Representative flow cytometry plots for patient 1 are shown in panels E to H. Compared to the empty vector control, RPS14 overexpression results in an increase in GlyA expression and a decrease in CD41 and CD11b.