RASA3 Is Essential for Early Mammalian Hematopoiesis and Modulates ERK Signaling in Terminal Erythropoiesis

The Ras GTPase-activating protein, RASA3, plays a critical non-redundant role in vertebrate erythropoiesis. The autosomal recessive scat mouse model carries a missense (G125V) mutation in Rasa3 that results in severe anemia, thrombocytopenia and leukopenia with markedly reduced survival. Terminal erythroid differentiation is dysregulated in scat bone marrow (BM) and spleen with delayed cell cycle progression, decreased apoptosis, and increased reactive oxygen species. To further study the role of RASA3 in hematopoiesis, we created a conditional Rasa3 allele. Germline null mice die at E12.5-13.5 of severe hemorrhage, pallor and reduced fetal liver size. Specific ablation of Rasa3 in hematopoietic stem and progenitor cells (HSPCs) and endothelial cells (ECs) driven by Vav-Cre and Tek-Cre recapitulates the null phenotype, highlighting the specific requirement of RASA3 in hematopoietic and vascular compartments. To study RASA3 deficiency specifically in terminal erythropoiesis, we utilized EpoR-Cre. Surprisingly, EpoR-Cre driven deletion does not produce an abnormal erythroid phenotype, and normal levels of RASA3 protein are seen on western blots of Ter119+ spleen and BM cells despite efficient deletion of the floxed Rasa3 allele, suggesting that RASA3 produced early in HSPCs is sufficient to maintain normal terminal erythroid differentiation. Indeed, deletion of Rasa3 in the entire hematopoietic system at 8 weeks of age via inducible Mx1-Cre results in death within 2 weeks of induction with profound anemia, thrombocytopenia, and leukopenia, recapitulating the scat phenotype. While HSC numbers are normal in the RASA3 depleted BM of this model, the frequency and absolute numbers of all progenitors are significantly decreased, and spleen stress erythropoiesis is significantly blunted.

Altered Ras activity is associated with delayed erythropoiesis. ERK signaling, a key downstream Ras effector, is hyperactivated 3-fold by western blotting in scat BM CD45-depleted cells vs. wild type (WT). To extend RASA3 studies to human cells, we utilized cord blood CD34+ cells in a three-phase erythroid differentiation culture system. Similar to the mouse models, shRNA knockdown of RASA3 leads to delayed erythroid differentiation at culture day 11 based on glycophorin A, Band 3 and a4-integrin expression levels. Dysregulated cell cycle progression with accumulation in the S-phase is also evident at day 12 of culture. Unlike the mouse models, however, ERK activation was decreased (67-72%) in serum-starved sorted proerythroblasts and later day 11 populations of RASA3 knockdown cells treated with erythropoietin (Epo), stem cell factor (SCF), or both. To elucidate additional molecular mechanisms downstream of RASA3, we performed RNAseq on total spleen and BM, as well as sorted stem and myeloid progenitors (SMPs) of scat mice and WT littermate controls. Principle component analysis confirms that genotype is the primary contributing factor to differential gene expression. In preliminary analyses, MAPK signaling, including negative regulators of ERK, phosphatidyl-inositol signaling, cell cycle regulation, and transcriptional regulation emerge from pathways and GO term analyses of differentially expressed genes.

In conclusion, we show that (1) absence of RASA3 results in embryonic lethality with evidence of severe HSPC/EC defects; (2) RASA3 is required for hematopoiesis starting at the HPC stage or earlier; (3) missense mutations (scat) or loss of RASA3 in adults cause pancytopenia and premature death; (4) mechanistically, aberrant ERK signaling and cell cycle defects contribute to the RASA3 deficient erythroid phenotype, while RNAseq studies provide further potential mechanistic targets.