A CRISPR Screen in Enucleated Red Blood Cells Identifies Critical Factors for Terminal Erythroid Differentiation

During terminal erythroid differentiation, committed precursor cells proliferate and differentiate down the erythroid lineage, yielding billions of new reticulocytes and erythrocytes each day. In mammals, the final transition from late-stage orthochromatic erythroblast to reticulocyte involves a striking morphologic change wherein the nucleus is expelled extracellularly, a process termed enucleation. Enucleation is known to involve transcription factors, microRNAs, and cytoskeletal-associated proteins, but a comprehensive understanding of how factors from the erythrocyte proteome may influence this complex process is lacking. To identify critical factors for erythroid enucleation, we designed and implemented a pooled CRISPR screen in cultured RBCs (cRBCs) differentiated from primary human bone marrow CD34+ hematopoietic stem cells (HSPCs). By employing a vector derived from CROPseq-Guide-Puro in which the hU6-sgRNA cassette is cloned into the lentiviral 3'LTR (Datlinger et al, Nat Methods, 2017), sgRNA sequences were incorporated into the puromycin-resistance mRNA transcripts exported to the cytoplasm, enabling their quantification in enucleated cells via RNA-seq. To implement the screen, we transduced primary human HSPCs with our custom lentivirus library on day 4 of differentiation, performed nucleofection with recombinant Cas9 the following day, and induced the cells to differentiate down the erythroid lineage. On day 17, we used a combination of FACS and RNA-seq to quantify the relative abundance of each sgRNA in the enucleated cRBCs relative to unsorted cRBCs. Analysis of three biological replicates identified a number of genes with underrepresented sgRNAs in the enucleated cRBCs, suggesting they target genes important for enucleation. Two of the top candidates were the Chloride Intracellular Channel 3 (CLIC-3) and Vesicle Associated Membrane Protein 8 (VAMP8), neither of which has a previously described role in terminal erythroid differentiation. Using immunoblotting and immunofluorescence assays, we demonstrated that both CLIC3 and VAMP8 are detectable throughout all stages of erythropoiesis, with expression decreasing over the course of differentiation. To validate these candidates, we generated knockdown (KD) cRBCs from primary human HSPCs. Monitoring of surface markers demonstrated reduced Band3 expression in the CLIC3 KD cRBCs by day 11, indicating disrupted erythroid differentiation. The day 11 cRBCs also had increased levels of P21 RNA and protein, suggesting cell cycle dysregulation. On day 13, the control cell population consisted primarily of polychromatic and orthochromatic erythroblasts, while the CLIC3 KD cells appeared to be earlier precursors, consistent with an erythropoiesis phenotype. Further supporting a role for CLIC3 in erythroid differentiation, by day 18 we observed that the CLIC3 KD population had only 30% enucleated cRBCs, as compared to 85% for the control cells. In contrast, the VAMP8 KD cRBCs demonstrated an accelerated rate of differentiation, with an earlier expression of Band3 on day 11 and an earlier loss of CD49d on day 13, as compared to control cRBCs. However, by the end of differentiation on day 18 only 38% of the cells were enucleated in the VAMP8 KD condition as compared to 85% of enucleated cells in the Mock transfected condition, suggesting a specific role for VAMP8 in enucleation. Our work highlights the existence of undiscovered factors important for terminal erythroid differentiation and enucleation, and the potential of forward genetic screening in enucleated RBCs to shed light on this final, relatively understudied step of erythropoiesis. Our ongoing efforts involve uncovering the precise molecular mechanisms underlying the novel roles of CLIC-3 and VAMP8 during erythropoiesis and enucleation.