Abstract
Myelodysplastic syndromes (MDS) are clonal hematopoietic stem cell disorders characterized by dysplastic hematopoiesis and peripheral blood cytopenias. Recently, somatic mutations affecting components of the spliceosomal machinery have been discovered in the majority of MDS patients. SF3B1 mutations are most frequent and strongly correlate with the presence of bone marrow ring sideroblasts and a favorable prognosis. SF3B1 mutations, including the K700E substitution which accounts for more than 50% of all mutations, are missense, heterozygous and cluster in a hotspot within the heat domain of the protein suggesting that they are gain-of-function variants.
The molecular effects of SF3B1 mutations and the mechanisms through which they drive clonal expansion and dyserythropoiesis remain obscure. Therefore, to assess their molecular and phenotypic consequences, we generated a mouse model carrying a conditional floxed knock-in allele (Sf3b1flox-K700E/+) by homologous recombination of JM8 murine embryonic stem cells. To induce expression of Sf3b1 K700E in adult hematopoietic stem and progenitor cells, Sf3b1flox-K700E/+/Mx1-Cre+ were injected with pIpC from 4-8 weeks of age. Here we report the initial characterization of these animals.
Monthly peripheral blood counts from mutants and wild-type (WT) littermates starting one month post-pIpC injection showed a reduction in hemoglobin levels (at 8 weeks WT=17g/dl mut=14.5g/dl, p<0.03). Additionally, flow cytometric analysis of bone marrow samples demonstrated a modest but consistent decrease in late erythroid progenitor cells (Ter119+ and CD71-/low). The myeloid compartment showed relative expansion of Gr1+/Mac1+ and Mac1+ cells whereas analysis of hematopoietic stem and progenitor cells (HSPCs) revealed a decrease in HSCs (% of total events WT=0.04%; Sf3b1flox-K700E/+=0.01%) in mutant mice. In competitive transplantation experiments into sub-lethally irradiated syngeneic recipients we observed a lower engraftment potential of Sf3b1flox-K700E Lin-ve HSPCs (CD45.2) compared to wild-type cells (CD45.1). Flow cytometric analysis of peripheral blood of recipient animals showed that Sf3b1flox-K700E cells contributed more to the myeloid lineage than wild-type cells (Sf3b1flox-K700E Mac1+/Gr1+ 8.95%; Mac1+ 15% vs WT Mac1+/Gr1+ 4.08%; Mac1+ 5.57%). At a median follow-up of 56 weeks, mutant animals did not show decreased survival or signs of illness as compared to WT controls.
Finally, as Sfb31 mutations are predicted to affect splicing of pre-mRNA and consequently alter the gene expression, we performed RNAseq analysis in unselected and Lin-ve bone-marrow cells from mutant and controls animals. Comparison between wt and mutant samples showed deregulated expression of genes implicated in human MDS (Mmp9, Puma, Bcl2l1). We then looked at the pattern of aberrant splicing promoted by Sf3b1flox-K700E, and found that mutant animals have an increased use of cryptic 3'' splice sites (ss) throughout their genome. We showed that the majority of these alternative 3' ss are novel and we characterized them as being located 15 to 24 nucleotides upstream from the canonical 3' ss and associated with sequence features including a shorter polypyrimidine tract and an enrichment of adenines -8 to -18 bases upstream of the cryptic 3' ss. Interestingly, similar features have been reported in human cancers with SF3B1 hotspot mutations. We predict that ~33% of the mRNAs affected by aberrant splicing will include an aberrant premature termination codon, promoting RNA degradation through nonsense-mediated decay.
In conclusion, our conditional Sf3b1K700E knock-in mouse is a faithful molecular model of the consequences of these mutations in the mouse hematopoietic system. The mild phenotype we observe in comparison to SF3B1-mutant human MDS may be explained by the requirement for additional mutations to progress to overt MDS and is more reminiscent of SF3B1-associated clonal hemopoiesis, relatively common phenomenon in elderly humans without overt hematological abnormalities.
Additionally, our initial characterization of novel splice sites preferentially recognised by the mutant Sf3b1 protein suggests that transcriptional consequences of the mutation may differ between species, dependant on the degree of conservation of the relevant intronic regions.
Seiler:H3 Biomedicine: Employment. Peng:H3 Biomedicine: Employment. Buonamici:H3 Biomedicine: Employment. Campbell:14M genomics: Other: Co-founder and consultant.
Author notes
Asterisk with author names denotes non-ASH members.