Abstract
Approximately 53% of patients with Diamond-Blackfan Anemia (DBA) have mutations in one of nine ribosomal protein components (RPS7, RPS10, RPS17, RPS19, RPS24, RPS26, RPL5, RPL11 and RPL35A), with RPS19 being the most commonly affected (25% of all patients). Mutations found in patients result in protein haploinsufficiency and subsequent ribosomal stress, which increases the expression and activity of p53, leading to a failure of HSC differentiation during erythropoiesis. While RPS19-/- mice are not viable, we have generated a model system using different shRNA constructs to generate varying degrees of RPS19 knockdown in cultured cells and animal models. Cultured cells (including CD34+ HSCs derived from fetal liver or cord blood) recapitulate features of DBA erythropoietic failure in vitro and in vivo, including p53 upregulation and defects in erythrocyte differentiation, while anemias and skeletal defects are observed in zebrafish models.
Using next-generation miRNA sequencing on the Illumina HiSeq2000 platform, we identified a number of miRNAs differentially expressed between human fetal liver CD34+ HSCs expressing normal or reduced RPS19. Preprocessing, alignment, and miRNA quantification were carried out using miRDeep2 [1] software. Differential expression analysis was performed with DESeq2 [2], which tests for differential expression based on a model using negative binomial distribution for count data from high-throughput sequencing assays. Candidates underwent a second round of screening based on robustness of expression differential (fold increase/decrease, variation of fold increase/decrease across replicates, and increase/decrease relative to RPS19 expression) and possible links to hematopoiesis and erythropoiesis, as reported in the scientific record. All three miR-34 isoforms (a, b and c) are robustly upregulated upon RPS19 reduction, with the extent of miR-34 upregulation being inversely proportional to RPS19 expression. When RPS19 protein expression is suppressed by approximately 50% (as observed in DBA patients), miR-34a, b and c isoforms are upregulated by 3.2, 2.2 and 2.6 fold respectively. In other systems, increased miR-34 results in modulation of a number of factors linked to erythropoiesis, including downregulation of c-Myc, Myb and NOTCH signaling. Genome-wide transcriptional analysis using next-generation sequencing on the Illumina HiSeq platform in CD34+ HSCs indicates c-Myc (4.2 fold), c-Myb (2.8 fold) and a number of NOTCH effectors [SIRT1 (1.8 fold) andHES1 (1.9 fold)] are indeed downregulated at the transcriptional level. Results were validated with qRT-PCR and western blot analysis, however, while transcriptional modulation fully accounts for c-Myc downregulation at the mRNA and protein level, we propose that c-Myb protein levels are modulated both transcriptionally and post-translationally, as protein expression is significantly further downregulated than mRNA message (6.3 fold). We propose that miR-34 upregulation in RPS19-depleted cells occurs through p53 and initiates genomic changes incompatible with erythropoiesis through downregulation of transcription factors c-Myc, c-Myb and the NOTCH signaling pathway.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.