Editing the HBG1/2 promoters or the BCL11A erythroid enhancer in CD34⁺ cells with Cas9 RNP robustly increases HbF expression but leads to different levels of transcriptomic perturbation in erythroid progeny in vitro. (A) Developmentally regulated hemoglobin switch during human ontogenesis. (B) BCL11A is a transcription factor that mediates the HbF to adult hemoglobin (HbA) switch shortly after birth by repressing the expression of γ-globin, a component of HbF, via the distal BCL11A binding sites at the γ-globin gene (HBG1/2) promoters. (C) Single-guide RNAs (sgRNAs) were designed to target the +58 GATA1 binding site at the BCL11A erythroid enhancer on chromosome 2 or the −110 distal BCL11A binding sites at the HBG1/2 promoters on chromosome 11. (D) Peripheral blood–mobilized CD34⁺ cells from healthy donors were prestimulated for 2 days and then electroporated with 1-8 μM of CRISPR-associated protein 9 (Cas9) complexed with guide RNA (gRNA) targeting either the HBG1/2 promoters (HBG1/2) or the BCL11A erythroid enhancer (BCL11A) at a gRNA-to-protein ratio of 4:1 using a Maxcyte GT electroporator, program HBM34-4. The editing rate, expressed as the frequency of sequencing reads with indels at the HBG1 promoter, HBG2 promoter, or hybrid HBG1-2 promoter (the result of simultaneous cleavage of HBG1 and HBG2 promoters and the deletion of the intervening sequence) was determined by next-generation sequencing, 1 day after electroporation. (E) CD34⁺ cells were placed in erythroid culture conditions for 18 days and the level of HbF expression was evaluated via reverse phase ultrahigh performance liquid chromatography in which individual globin chains (α, β, Aγ, and Gγ) were separated. HbF level was calculated as (Aγ + Gγ)/(β + Aγ + Gγ) × 100%. Data from CD34⁺ cells electroporated with 4 μM or 8 μM RNP are shown, with maximum levels of editing (typically >95%) achieved in erythroid cells. (F) Day-11 erythroid cells in triplicates from CD34⁺ cells transfected with 8 μM RNP were collected and the transcriptome profiled via RNA sequencing. High levels of editing were achieved in erythroid cells (mean ± standard deviation: HBG1/2, 98.68% ± 0.09%; BCL11A, 99.22% ± 0.08%). Transcripts expressed at levels of more than or equal to twofold difference from control samples, and FDR of <0.05 were identified (red, blue [BCL11A], and orange [HBG1/2] dots). (G) Differentially expressed genes (DEGs) involved in the cell processes of apoptosis, differentiation, proliferation, or erythropoiesis in BCL11A-edited samples are listed. For panels D-E, each circle represents an independent sample, and the horizontal bar represents the mean value of the respective treatment group. n = 4 to 7 for panel D and n = 6 to 11 for panel E. Mobilized peripheral blood CD34⁺ cells from multiple donors were used; n = 9 for HBG1/2 and n = 6 for BCL11A. Ordinary 1-way analysis of variance (ANOVA) with Tukey multiple comparisons test was performed to evaluate the statistical significance of the mean values for panel D. ∗P < .05; ∗∗P < .001; ∗∗∗∗P < .0001. BGLT3, β-globin locus transcript 3; BMPR1A, bone morphogenetic protein receptor type 1A gene; CARD9, caspase recruitment domain family member 9 gene; chr, chromosome; ex, exon; FC, fold change; FDR, false discovery rate; FOXC1, forkhead box C1 gene; FZD4, frizzled class receptor 4 gene; GATA5, GATA-binding protein 5 gene; G0S2, G0/G1 switch 2 gene; HBB, β-globin gene; HBD, δ-globin gene; HBE, ε-globin gene; HES4, hes family BHLH transcription factor 4; HS, hypersensitive site; IGF2, insulin-like growth factor 2 gene; IGFL2, IGF-like family member 2 gene; IRS4, insulin receptor substrate 4 gene; JAG1, jagged canonical notch ligand 1 gene; KLF15, KLF transcription factor 15 gene; LCR, locus control region; PAM, protospacer adjacent motif; PDE3A, phosphodiesterase 3A gene; PDGFA, platelet-derived growth factor subunit A gene; SMAD3, SMAD family member 3 gene; TFCP2L1, transcription factor CP2-like 1 gene; TNFRSF25; TNF receptor superfamily member 25 gene; TP53I11; tumor protein p53 inducible protein 11 gene.