Micro RNA (miRNA)-mediated gene silencing, largely mediated by the Argonaute (AGO) family proteins, is a post-transcriptional gene expression control mechanism that has been shown to regulate hematopoietic stem and progenitor cells (HSPCs) quiescence, self-renewal, proliferation, and differentiation. Interestingly, only the function of AGO2 in hematopoiesis has been investigated. O'Carroll et al. (2007) showed that AGO2 knockout in mice bone marrow cells interferes with B220low CD43- IgM-pre-B cells and peripheral B cell differentiation and impairs Ter119high, CD71high erythroid precursors maturation. However, the functional significance of other AGO proteins in the regulation of stemness and lineage commitment remains unclear. AGO submembers, AGO1-4 in humans, are traditionally believed to act redundantly in their function. However, our previous proteomic analysis from sorted populations of the human hematopoietic hierarchy shows each sub-member is differentially expressed during HSPCs development, suggesting each sub-member may have a specialized function in hematopoiesis. Here, we conducted CRISPR-Cas9 mediated knockout of AGO1-4 in human cord blood derived long-term (LT-) and short-term hematopoietic stem cells (ST-HSCs) and investigated the impact of the loss of function of individual AGOs in vitro and in vivo in xenograft assays.

From the in vitro experiment, we cultured CRISPR-edited LT- or ST-HSCs in a single cell manner on 96-well plates pre-cultured with murine MS5 stroma cells in erythro-myeloid differentiation condition. The colony-forming capacity and lineage commitment of each individual HSC is assessed on day 17 of the culture. Initial data showed that AGO1, AGO2 and AGO3 knockout decreased the colony formation efficacy of both LT- and ST-HSCs, suggesting AGO1, AGO2 and AGO3 are involved in LT- and ST-HSCs proliferation or survival. As for lineage output, AGO1 knockout increases CD56+ natural killer cell commitment in LT-HSCs and erythroid differentiation in ST-HSCs; AGO2 knockout increases erythroid differentiation in both LT- and ST-HSCs and decreases myeloid differentiation in ST-HSCs; while AGO4 knockout seems to decrease erythroid output.

For the in vivo experiment, we xenotransplanted AGO1 and AGO2 knockout LT-HSCs in irradiated immunodeficient NSG mice and assessed the change in LT-HSCs engraftment level and lineage differentiation profile at 12- and 24-week time points. We found that AGO2 knockout increased CD45+ engraftment at both 12- and 24-weeks. Aligning with our in vitro data, AGO2 knockout increases GlyA+ erythroid cells at 12- and 24-weeks. The increase in GlyA+ erythroid cells is a consequence of the 2-fold increase in GlyA+ CD71+ erythroid precursor cells, recapitulating previous findings that AGO2 knockout in mice impairs CD71high erythroid precursor maturation leading to the accumulation of undifferentiated CD71+ erythroid precursors (O'Carroll et al., 2007). Accumulation of early progenitors of the erythroid lineage, including the common myeloid progenitors (CMPs) and myelo-erythroid progenitor (MEPs) were observed, as well as their progeny including CD33+ myeloid and CD41+ megakaryocytes. For the myeloid lineage, AGO2 knockout shifts myeloid differentiation toward CD66b+ granulocytes from CD14+ monocytes. For lymphoid, AGO2 knockout decreases CD19+ CD10- CD20+ mature B-lymphoid cells, which again aligns with previous AGO2 knockout mice results. On the other hand, AGO1 knockout LT-HSCs share some similar phenotype with AGO2 knockout LT-HSCs, where AGO1 knockout increases CD71+ erythroid precursors. However, AGO1 knockout in LT-HSCs also results in unique phenotypes, with a decrease in neutrophil formation and an increase in CD4+ CD8+ T progenitor cells are observed. AGO3 and AGO4 knockout experiments are in progress.

In summary, our AGO2 knockout experiments recapitulate the reported results from murine studies but also illustrate a more complete role of AGO2 in hematopoietic lineage differentiation. Moreover, AGO knockout experiments of individual submembers are revealing novel insights into their role in the regulation of stemness and lineage commitment of LT-HSCs and ST-HSCs. These data point to a unique role of different AGO isoforms in lineage commitment in human HSCs and argue against redundant functioning.


Dick:Bristol-Myers Squibb/Celgene: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.