Editing the HBG1/2 promoters is compatible with multilineage engraftment and robust HbF induction, whereas editing the BCL11A erythroid enhancer affects erythroid production in a xenotransplantation model. Mobilized healthy donor CD34⁺ cells were electroporated with 1 to 8 μM of ribonucleoprotein complexed at a 2:1 gRNA-to-protein ratio using a Maxcyte GT electroporator, program HBM34-4. Cells were cryopreserved 1 day after electroporation. Cells with comparable levels of editing were later thawed and infused intravenously into NBSGW mice at 1 × 10⁶ cells per mouse. (A) At 18 weeks after infusion, animals were euthanized to determine the levels of human engraftment in the BM by flow cytometry. Chimerism was calculated as hCD45⁺/hCD45⁺+mCD45⁺; human HSPC frequency was calculated as hCD34⁺/hCD45⁺; human B-cell frequency was calculated as hCD19⁺/hCD45⁺; human neutrophil frequency was calculated as hCD15⁺/hCD45⁺; and human erythroid frequency was calculated as hCD235a⁺/total viable cells. Human T-cell engraftment was at the background level and therefore not reported. (B) Erythroid output (hCD235a⁺ per total viable cells) was adjusted to HSPC frequency (hCD34⁺ per total viable cells) in corresponding animals to evaluate the ability of HSPCs to generate erythroid cells in vivo. (C) HbF expression was determined by reverse phase ultrahigh performance liquid chromatography in human erythroid cells isolated by flow-activated cell sorting from mouse BM. (D) Editing levels of infused CD34⁺ cells (before) and unfractionated BM were determined by next-generation sequencing. (E) Human HSPCs (hLin⁻hCD34⁺hCD45⁺), B cells (hCD19⁺hCD45⁺), neutrophils (hCD15⁺hCD45⁺), and erythroid cells (hCD235a⁺) were flow sorted and analyzed for editing levels by next-generation sequencing. (F) Editing data of flow-sorted human cells from 3 independent in vivo studies were compiled, and correlation analyses were performed for B cells, neutrophils, or erythroid cells against HSPCs. The black angled line represents the identity line (y = x). (G) Editing data of flow-sorted HSPCs from BM, and erythroid output from 3 independent in vivo studies were compiled, and correlation analysis was performed. To adjust for the differential erythroid output observed in each study, erythroid output was normalized to the output observed in their respective control samples. Each circle represents 1 sample from 1 animal. n = 7 to 8 per treatment group for panels A-E; n = 33 to 39 for panel F; and n = 41 to 52 for panel G including controls. For panels E-G, samples with reads of <5000 mapped to the target were excluded. For panel G, samples with human chimerism levels of <50% were excluded. For panels A-C, ordinary 1-way ANOVA with Tukey multiple comparisons test was performed. For panel E, mixed-effects analysis with Tukey multiple comparisons test was performed to compare indel levels of samples sorted from the same animals. For panel G, simple linear regression analysis was performed, and the slope was tested for nonzero significance. ∗P < .05; ∗∗P < .001; ∗∗∗∗P < .0001.