Base editing to correct inborn errors of immunity

Abstract

Described herein are compositions and methods for highly efficient and specific base-editing to correct two distinct X-linked chronic granulomatous disease (X-CGD) mutations, CYBB c.676C>T and c.1075G>A.

Description

[0001]Attorney Docket No.29539-0835WO1 / MGH 2024-374 BASE EDITING TO CORRECT INBORN ERRORS OF IMMUNITY CLAIM OF PRIORITY This application claims the benefit of U.S. Provisional Patent Application Serial No.63 / 707,548, filed on October 15, 2024. The entire contents of the foregoing are hereby incorporated by reference. FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT This invention was made with Government support under Grant Nos. HL142494 and CA281401, and Intramural Project Nos. Al00644 and Al00988, all awarded by the National Institutes of Health. The Government has certain rights in the invention. TECHNICAL FIELD Described herein are compositions and methods for highly efficient and specific base-editing to correct two distinct X-linked chronic granulomatous disease (X-CGD) mutations, CYBB c.676C>T and c.1075G>A. BACKGROUND The continued expansion of genetic diagnoses for inherited disorders has revealed a great opportunity to explore new classes of genetic therapies as definitive treatments. Hematopoietic stem cell transplants are potential cures for certain classes of diseases but are difficult to scale due to graft versus host disease and a lack of HLA-matched donors. Gene therapies based on the modification of autologous hematopoietic stem cell and progenitor cells (HSPCs) overcomes some of these challenges. Much progress has improved the safety and efficacy of viral vector- mediated genetic modification of patient HSPCs. Current self-inactivating lentivectors have demonstrated clinical benefit in multiple inborn errors of immunity (IEI), including Chronic Granulomatous Disease, X-linked Severe Combined Immunodeficiency (X-SCID), ADA-deficient SCID, and Wiskott Aldrich Syndrome (1-4). However, there remain safety concerns related to vector insertional oncogenesis in Cerebral Adrenoleukodystrophy (5) and HMGA2+ clonal expansion in X-SCID (6) and ^-Thalassemia (7), and limited development of therapeutic lentivectors for only a Attorney Docket No.29539-0835WO1 / MGH 2024-374 few IEIs. Thus, there is a major unmet need for efficient and safe genetic therapies for IEIs. Recent progress in the development of genome editing technologies provides hope to solve these key challenges. Next-generation genome editing approaches permit user-specifiable genetic alterations to functionally correct patient mutations. By enabling the correction of the underlying genetic mutations, rather than overexpressing a corrective transgene, genome editing approaches offer an improved safety profile with broad applicability for many IEIs. Chronic granulomatous disease (CGD) is an inherited disorder that is typically caused by a defective phagocyte NADPH oxidase complex that produces bactericidal reactive oxidative species for host defense. CGD patients experience recurrent invasive infections, hyperinflammation and inflammatory bowel disease (8-10). Despite lifelong antibiotic prophylaxis, CGD patients experience increased morbidity and early mortality with infection rates of 0.3 / year despite anti-microbial prophylaxis (9). Ex vivo gene therapy for X-linked CGD (X-CGD) to restore canonical gp91phoxprotein expression using lentivector-transduced autologous HSPCs has treated 19 patients to date, three of whom have lost their grafts for reasons yet unclear (2). Peripheral blood granulocytes from treated patients express gp91phoxat approximately 1 / 3 of normal donor levels due to a relatively weak chimeric promoter (2, 11). Permanent semi-random vector insertion, the use of weak exogenous promoter to reduce risks of transactivating nearby oncogenes, and the lack of physiological regulation provided rationale to explore other approaches including targeted genome editing, which should in principle address these issues. SUMMARY Described herein are methods using ABE8e-SpRY and ABE8e-WT for ABE- mediated correction of several CYBB mutations for treatment of X-CGD, which achieved high levels of on-target editing that resulted in efficient and precise mutation correction in HSPCs, leading to phenotypic restoration in ex vivo edited primary cells. High-fidelity SpRY enzymes (28)(29) were used to improve the safety profile by reducing off-target and bystander edits observed with conventional SpRY. The flexible ABE8e-SpRY base editor is a promising and broadly applicable approach for ex vivo precision treatment of IEIs that are caused by single nucleotide mutations. Provided herein are compositions comprising a recombinant SpCas9 adenine base editor protein and a gRNA targeting CYBB, preferably as a ribonucleoprotein Attorney Docket No.29539-0835WO1 / MGH 2024-374 complex, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. Additionally provided herein are compositions comprising a nucleic acid sequence encoding an SpCas9 adenine base editor protein, and a gRNA targeting CYBB, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. In some embodiments, the gRNA targets the BE to CYBB mutation c.676C>T and comprises A3, A4, or A5 gRNA (e.g., one of SEQ ID NOs:1-3), and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e-SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (e.g., SEQ ID NO:4) with ABE8e- SpG or ABE8e-NRRH. In some embodiments, the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6, A7, A8 (e.g., one of SEQ ID NOs:5, 4, or 6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (e.g., SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY. In some embodiments, the recombinant SpCas9 adenine base editor protein and gRNA targeting CYBB are in a carrier, e.g., a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle. Also provided herein are compositions comprising a nucleic acid sequence encoding an SpCas9 adenine base editor protein and a nucleic acid sequence encoding a gRNA targeting CYBB, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. In some embodiments, the nucleic acid sequences encoding the base editor protein and / or gRNA are in a vector, e.g., a viral vector or a plasmid, and / or in a carrier, e.g., a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle. Further, provided herein are methods of editing a cell comprising a mutation in an allele of CYBB. The methods comprise contacting the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. Also provided herein are isolated cells from a subject who has a disease caused by a mutation in CYBB, wherein the mutation has been corrected by contacting Attorney Docket No.29539-0835WO1 / MGH 2024-374 the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. In some embodiments, the cell is a hematopoietic stem cell and progenitor cell (HSPC). Additionally, provided herein are methods of treating a subject who has a disease caused by a mutation in CYBB. The methods comprise administering to the subject a therapeutically effective amount of an isolated autologous cell, wherein the mutation in the cell has been corrected by contacting the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. In some embodiments, the cell is a hematopoietic stem cell and progenitor cell (HSPC). Further, provided herein are methods of treating a subject who has a disease caused by a mutation in CYBB. The methods comprise administering to the subject a therapeutically effective amount of a nucleic acid encoding an SpCas9 adenine base editor protein and a gRNA targeting CYBB, or a recombinant SpCas9 adenine base editor protein and a gRNA targeting CYBB, preferably as a ribonucleoprotein complex, as described herein, preferably wherein the gRNA comprises a sequence shown in Table 2 or 3, e.g., of one of SEQ ID NOs:1-49. In some embodiments, the nucleic acid sequences encoding the base editor protein and / or gRNA are in a vector, e.g., a viral vector or a plasmid, and / or in a carrier, e.g., a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle. In some embodiments, the nucleic acid sequence encoding the base editor protein is an mRNA, and the method comprises administering the mRNA and gRNA in an LNP. In some embodiments, the recombinant SpCas9 adenine base editor protein and gRNA targeting CYBB are administered in a carrier, e.g., a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle. In some embodiments described herein, the gRNA targets the BE to CYBB mutation c.676C>T and comprises A3, A4, or A5 gRNA (e.g., one of SEQ ID NOs:1- 3), and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably Attorney Docket No.29539-0835WO1 / MGH 2024-374 ABE8e-SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (e.g., SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH. In some embodiments described herein, the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6, A7, or A8 (e.g., one of SEQ ID NOs:5, 4, or 6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT-HF1, or A5 (e.g., SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Methods and materials are described herein for use in the present invention; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting. All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control. Other features and advantages of the invention will be apparent from the following detailed description and figures, and from the claims. DESCRIPTION OF DRAWINGS FIGs.1A-I. PAMless base editing to model and correct the CYBB c.676C>T mutation. (a) Schematic of cytosine base editors (CBEs)(16, 50) that can be utilized to programmably install C-to-T edits. nSpCas9, nickase SpCas9(D10A); UGI, uracil glycosylase inhibitor; gRNA, single guide RNA. (b) Schematic of a region of the human CYBB gene near amino acid position R226. CBE target sites were designed to create the c.676C>T mutation (causing p.R226X); the position of the target cytosine in the target site protospacer is shown, along with the PAM (protospacer-adjacent motif). (c) Representative Sanger sequencing traces of control, heterozygous, and homozygous HEK 293T cell lines bearing the CYBB c.676C>T mutation. (d) Adenine base editors (ABEs)(14, 67) can install A-to-G edits. nSpRY, nickase SpRY(D10A), an engineered SpCas9 enzyme that can recognize an expanded range of protospacer adjacent motifs (PAMs)(22). (e) Schematic of the CYBB c.676C>T genomic locus, with gRNAs designed for ABEs to reverse the mutation and correct the c.676C>T substitution. The position of the target adenine in the target site Attorney Docket No.29539-0835WO1 / MGH 2024-374 protospacer is shown, along with the PAM. (f) A-to-G editing efficiency of the target adenine in a clonal heterozygous HEK 293T cell bearing the CYBB c.676C>T mutation, when using ABE8e-SpRY and various gRNAs with position of the target base in the spacer indicated. Mean and standard deviation shown with dots for individual datapoints (n = 2 or 3 independent biological replicates). (g) Schematic of the intended target base where editing corrects CYBB c.676C>T, as well as other proximal adenine and cytosine bases that could be subject to unwanted bystander edits. (h) Analysis of bystander editing at two adenine and cytosine bases proximal to the intended edit. Mean and standard deviation shown with dots for individual datapoints (n = 2 or 3 independent biological replicates). (i) A-to-G editing efficiency of the target c.676C>T adenine in heterozygous HEK 293T CYBB c.676C>T cells, using fidelity-enhanced ABEs(28, 29). Mean and standard deviation shown with dots for individual datapoints (n = 2 or 3 independent biological replicates). FIGs.2A-J. Optimization of base editing in healthy donor CD34+cells using a surrogate gRNA. (A) Targeted DNA sequence surrounding the CYBB c.676 locus. The protospacer associated motif (PAM) and target nucleotide are indicated for the surrogate CYBB i6 (-2) gRNA. (B) Schematic of processing CD34+hematopoietic stem / progenitor cells from patients and healthy donor controls. Overview of cell culture and editing conditions after patient apheresis. Conc= concentration; EP = editing procedure. (C) Percentages of edited alleles in HD CD34+cells electroporated with ABE8e-SpRY mRNA and the surrogate CYBB i6 (-2) gRNA at day 1 (D1) versus day 2 (D2) after cell culture (D1: mean = 30.97, SD = 5.69, N = 6; D2: mean = 42.24, SD =7.27, N = 5). Mann-Whitney test; *P < 0.05. (D) Percentage of edited alleles and the specific EP and culture conditions of HD CD34+cells. (E) Representative dot plot of gp91phox+expression in myeloid differentiated CD34+cells after 14 days in culture. (F-J) Analysis of cells in mice injected with non-edited healthy donor patient cells (HD: N = 4), base-edited once healthy donor cells (1xEP: N = 15) and base-edited twice healthy donor cells (2xEP: N = 8). (F) Change in CD15+gp91phox+expression in mice blood at weeks 12 and 16 after injection of HSPCs into mice. All data points are gated on CD45+. (G) Engraftment rates (CD45+) in bone marrow and spleen at mouse endpoint. (H) Percentage of neutrophils (CD15+), gp91phox+and DHR expression in human isolated (hCD45+) mouse bone marrow cells after ex-vivo myeloid differentiation for 14 days. Mann-Whitney test; ***P < 0.001, **P < 0.01. (I-J) Attorney Docket No.29539-0835WO1 / MGH 2024-374 Percentages of T (CD3+), myeloid (CD15+), and B (CD19+) cells in the (J) bone marrow and (I) spleen at endpoint after gating on CD45+. Data is shown as mean ± standard deviation. Each symbol indicates a single animal. FIGs.3A-E. Base editing to correct CYBB c.676 C>T CGD patient CD34+cells. (A) Percent of wild-type (WT) sequence (seq) in CGD patient HSPCs with the CYBB c.676C>T mutation treated with ABE8e-SpRY mRNA and either the A4 or A5 gRNAs once (1x EP) or twice (2x EP) (A41xEP: mean = 42.3, SD = 5.0 , N = 6; A4 2xEP: mean = 52.5 , SD = 5.5, N = 12; A51xEP: mean = 55.9 , SD = 8.3, N = 4; A5 2xEP: mean = 69.1 , SD = 6.1, N = 3; BE 1xEP: mean = 84.4, SD = `17.2, N = 7). Mann-Whitney test; **P < 0.01, ns, not significant, P ≥ 0.05. (B) Percentage of wildtype (WT) nucleotides after treatment of CGD CYBB c.676 patient cells with ABE8e-SpRY mRNA and the A5 gRNA at 6 different RNAse inhibitor concentrations (0, 10, 20, 30, 40, 60 x103Units / mL) compared to naïve CGD patient and healthy donor (HD) HSPCs. (C-D) Representative dot plots of (C) gp91phox+expression and (D) DHR+expression in myeloid differentiated CD34+cells after 14 days in culture. HD, CGD naïve, and CGD cells treated with ABE8e-SpRY mRNA and the A5 gRNA (CGD BE) were compared. (E) Percentage of CD15+and gp91phox+expression in myeloid differentiated CD34+cells after 14 days in culture. (Naïve: N = 4; 1xEP BE: N = 9; 2xEP BE: N = 5; HD: N = 5). Mann-Whitney test; ***P < 0.001, *P < 0.05. Data is shown as mean ± standard deviation. Each symbol indicates a single independent experiment. FIGs.4A-H. Base editing of CYBB c.1075G>A CGD patient CD34+cells. (A) DNA sequence of CYBB c.1075 G>A (p.Asp360Gly) mutation locus and single guides targeting the guanine (G) mutation numbered according to position (A5 to A9) in relation to the respective number of nucleotides upstream from the protospacer associated motifs (PAM). (B) Percentages of targeted correction in HSPCs treated with 6 different gRNA and ABE combinations, from experiments using either a single or double electroporation protocol (1x and 2x, respectively). Percentage of edited alleles were compared to healthy donor (HD) and non-edited CGD patient (naïve) HSPCs. (C) Representative dot plot of gp91phox+expression in myeloid differentiated healthy donor CD34+cells. Healthy donor (HD), CGD naïve, and base-edited CGD (CGD BE) were compared. (D) Percentage of CD15+and gp91phox+expression in myeloid differentiated CD34+cells after 19 days (d19) and 33 days (d33) were Attorney Docket No.29539-0835WO1 / MGH 2024-374 compared between healthy donor (HD), CGD base-edited (A5 SpG, A6 WT, A7 WT, A8 WT, A9 SpRY, A9 SpRY HiFi, A7 WT 2xEP), and CGD naïve. (E) DNA sequence displaying percentage of corrected allele in CGD patient CD34+cells after base- editing. Arrow indicates target nucleotide. (F-G) Percentage of stem cell subpopulations 2 days after base-editing. (F) Percentage of CD34+CD38- and CD34+CD38- CD90+CD133+HSPC subpopulations in base-edited (CYBBc.1075G>A, ABE8e SpWT / A7) once (1xEP) and base-edited twice (2xEP) CGD patient cells. (1xEP: N = 8; 2xEP: N = 7). (G) Percentage of edited alleles in healthy donor HSPC subpopulations after 1EP (green) versus 2EP (blue). Healthy donor cells were edited using ABE8e-SpRY-HF1 and gRNA i6(-2). (H) Percentage of normal (wildtype) nucleotide (NT) after base-editing repair of CYBB c.676C<T mutation (with gRNA A4 or A5) and CYBB c.1075 mutation (gRNA A8) in CGD patient HSPCs (c.676 A4: mean = 49.5, SD = 8.8, N = 19; c.676 A5: mean = 51.7, SD = 12.5, N = 10; c.1075 BE: mean = 84.4, SD = 17.2, N = 7). Mann-Whitney test; ***P < 0.001; **P < 0.01. Data is shown as mean ± standard deviation. FIGs.5A-K. Assessment of efficacy of base editing in HSPCs in-vivo. (A) Schematic for xenotransplant studies of adult immunodeficient mice (NSG) by IV injection of HSPCs for in vivo analysis of genomic and functional correction. BM = bone marrow; D = day; Pb = peripheral blood; Wk = week. (B) Representative dot plot of gp91phox+expression in mice blood 16 weeks after engraftment of human cells. Gp91phox+expression is first gated on CD45+and compared between healthy donor (HD), CGD patient non-edited (CGD naïve), and CGD patient base-edited (CGD BE). (C-K) Analyses of mice transplanted with CGD patient (naïve, red), one time base- edited (1xEP, blue), two time base-edited (2xEP, purple) and healthy donor (HD, green) HSPCs. (C.) Engraftment rates (hCD45+) in peripheral blood of transplanted mice at weeks 12, 14-16, and 20-23 after transplant. (D) Expression of gp91phox+in peripheral blood at weeks 8, 12, and 20-26 after transplant. (E) DHR expression in peripheral blood at week 20 after transplant. (F) Human cell engraftment (CD45+) in bone marrow at mouse endpoint and (G) expression of myeloid (CD15+), gp91phox+, and DHR+from isolated mouse BM human (CD45+) cells after ex-vivo differentiation for 14 days in culture. (H-I) percentages of T (CD3+), B (CD19+), and myeloid (CD15+) cells in the (H) peripheral blood and (I) bone marrow at mouse endpoint after gating on CD45+cells. (J) Expression of myeloid (CD15+) and gp91phox+cells in Attorney Docket No.29539-0835WO1 / MGH 2024-374 naïve, 1xEP BE, 2xEP BE, and HD HSPCs after 14 days of in-vitro myeloid differentiation in culture. Mann-Whitney test; **P < 0.01. (K) Expression of gp91phoxand DHR in isolated human (CD45+) cells from mouse BM after ex-vivo myeloid differentiation for 14 days in culture. Data is shown as mean ± standard deviation. Each symbol indicates a single animal. FIGs.6A-D. Correction in CGD patient HSPCs using high fidelity base editor enzymes. (A) Percentages of edited alleles in CGD patient c.676 HSPCs treated with 1 wildtype (A5 / SpRY) and 2 high-fidelity (A5 / SpRY HF 434, A5 / SpRY HF 458) guide and base-editor combinations (A5 / SpRY: mean = 42.8 , SD = 9.7, N = 4; A5 / SpRY HF 434: mean = 31.7 , SD = 5.5, N = 3; A5 / SpRY HF 458: mean = 32.7 , SD = 5.6, N = 4). Mann-Whitney test; ns, P ≥ 0.05. (B) Percentage of edited alleles determined by high throughput sequencing using different guides with varying concentrations (SpRY 0.5 uL, SpRY 1.0 uL, SpRY HF 4341.0 uL, SpRY HF 4341.5 uL, SpRY HF 4581.0 uL, SpRY HF 4581.5 uL). (C) Percentage of corrected wildtype (WT) nucleotides in human (CD45+) cells isolated from mouse spleen and bone marrow cells. Mice were transplanted with BE CGD c.676 patient HSPCs (A5 SpRY, A5 SpRY HiFi 434, A5 SpRY HiFI 458) and naïve CGD c.676 HSPCs after 4 days and 12 days in culture. (D) Percentage of edited alleles in CGD patient c.1075 HSPCs treated with 1 wildtype (WT) and 1 wildtype high fidelity (WT HF) base- editor. Data is shown as mean ± standard deviation. Each symbol indicates a single independent experiment. FIGs.7A-D. Analysis of on- and off-target base editing in HSPCs. (A) Manhattan plot results for nomination of potential off-target sites (OTs) by CHANGE- seq, using genomic DNA from two different X-CGD c.676C>T patient donors. The DNA samples were treated with either SpRY or SpRY-HF1 nuclease along with gRNA A5. (B) Analysis of on- and off-target base editing in X-CGD c.676C>T HSCs that were untreated (naïve) or treated with ABE8e-SpRY or ABE8e-SpRY-HiFi mRNA and a synthetic A5 gRNA. Genomic DNA was subjected to rhAmpSeq for on- target site and the top 50 CHANGE-seq nominated off-target sites, with data analysis via CRISPResso2; base editing efficiency plotted for only the most edited based for each target site, either an adenine or cytosine typically in the middle of the edit window; off-targets #11 and #23 omitted due to low read counts; the dashed line represents 0.5% base editing. (C) Whole exome sequencing of CGD c.676 BE patient Attorney Docket No.29539-0835WO1 / MGH 2024-374 HSPCs detected >80% correction of the target base (A>G on the opposite strand) and ~20-30% editing of a bystander base (C>G on the opposite strand). (D) Schematic illustrating the impact of the bystander edit of the exon 7 splice acceptor that can cause exon skipping in CYBB transcripts (e.g. exon 5 to exon and exon 6 to exon 8) uncovered by RNA sequencing of mutation (CYBB c.676C>T) specific cells edited with ABE8e-SpRY-HiFi and gRNA A5. FIGs.8A-E. Analysis of genome-scale consequences in base edited HSPCs. (A) Chromosomal arrangement of CGD patient HSPCs after base editor treatment, determined by karyotyping. (B) Optical genome mapping (Bionano) revealed all structural variants in CYBB c.676C>T patient cells treated with ABE8e-SpRY and gRNA A5 (left panel) and after filtering out variants also observed in untreated naïve sample (right panel), leaving only one SV (see panel C). (C) Manual inspection of the SV in ABE8e-SpRY treated cells revealed that it was also observed at a low level in naïve cells. (D) Representative images of comet assay to detect chromosomal abberations. Presence of comet tails in CGD patient HSPCs (light green) after treated with different conditions (naive, Cas9 / AAV, BE) determined by single cell electrophoresis comet assay. (E) Percentage of CGD patient HSPCs displaying comets were compared after different treatment conditions, including non-edited (Naïve), Puromycin treatment (positive control), Cas9 nuclease with a gRNA and an AAV HDR donor, Cas9 nuclease with a gRNA and an AAV HDR donor + hu-genetic suppressor element (GSE)-56 (42) (49), base edited with ABE8e-SpRY and gRNA A5 with either a single or double electroporation of mRNA. Comets were measured 2, 3, and 5 days (“D”) after electroporation. FIGs.9A-E. Use of cytosine base editors to create a CYBB- c.676C>T / p.R226X cell line. (a) Workflow for assessing the efficiency of different CBE constructs and gRNAs to generate the R226X mutation in HEK 293T cells in bulk transfection experiments. The CBE construct BE4max is comprised of an engineered APOBEC1 deaminase domain(50); SpG and SpRY are engineered SpCas9 enzymes that can recognize an expanded range of protospacer adjacent motifs (PAMs)(22). (b) C-to-T editing efficiency of the target c.676C>T cytosine to create the CYBB-R226X mutation, as assessed by targeted sequencing. Mean and standard deviation shown, with dots for individual datapoints for n = 4 replicates. (c) Exemplary next-generation sequencing result for a bulk population of HEK 293T Attorney Docket No.29539-0835WO1 / MGH 2024-374 cells that were transfected with BE4-SpRY and gRNA C6 (analyzed using CRISPResso2; (56)). The desired edit to create CYBB-R226X was achieved in approximately 15.41% of genomes. (d) Workflow for generating a clonal HEK 293T cell line harboring the R226X mutation. (e) Genotype of HEK 293T clones bearing the homozygous or heterozygous CYBB-R226X mutations (left and right panels, respectively). Genotype determined via next-generation sequencing with data analyzed using CRISPResso2. FIG.10. Use of adenine base editors to correct CYBB-c.676C>T. Workflow for assessing the on-target editing efficiencies of different ABE8e constructs and gRNAs to correct the CYBB-c.676C>T mutation in a homozygous clonal HEK 293T cell line in bulk transfection experiments. FIGs.11A-B. Bystander editing when using ABEs to correct CYBB- c.676C>T. (a) Next-generation sequencing results (analyzed using CRISPResso2; (56) for a bulk population of CYBB-c.676C>T HEK 293T cells that were transfected with ABE8e-SpRY and gRNA A5. Schematic of amplicon (left panel) and editing results (right panel). (b) Analysis of bystander editing at two adenine and cytosine bases proximal to the intended CYBB-c.676C>T edit when using ABE8e-SpRY-HF1 or ABE8e-SpRY-HiFi (left and right panels, respectively). Mean and standard deviation shown with dots for individual datapoints (n = 2 or 3 independent biological replicates). FIG.12. Comparison of SpCas9 PAM variant ABEs to correct CYBB c.676C>T when using gRNA A7. A-to-G or C-to-T editing efficiencies of the target adenine and proximal bystander adenines or cytosines (right panel). Experiments were performed in a clonal heterozygous HEK 293T cell bearing the CYBB c.676C>T mutation, when using different SpCas9 PAM variant ABEs with gRNA A7 (left panel). Mean and standard deviation shown with dots for individual datapoints (n = 2 or 3 independent biological replicates). FIGs.13A-B. Summary of on- and off-target base editing. A-B, The on- target site and top 50 CHANGE-seq nominated off-target sites were subjected to targeted amplicon sequencing via rhAmpSeq, utilizing genomic DNA from X-CGD c.676C>T HSCs either untreated (Naïve) or treated with ABE8e-SpRY or ABE8e- SpRY-HiFi mRNA along with a synthetic gRNA A5. Data from the rhAmpSeq output was analyzed using CRISPResso2; the base editing efficiencies were plotted for all Attorney Docket No.29539-0835WO1 / MGH 2024-374 adenine or cytosine bases across a wide edit window (defined as bases 1 through 12 in the target site spacer, counting from the PAM distal end of the spacer); off-target sites 11 and 23 failed to yield results due to low sequencing read counts. FIGs.14A-B. Whole exome analysis of base editor treated cells. (A) Method flow chart for analysis of whole exome sequencing experiments using genomic DNA from ABE8e-SpRY and gRNA A5 treated CYBB c.676C>T cells from two separate X-CGD patients. (B) Venn diagrams to plot the variants identified by whole exome sequencing in samples from two CYBB c.676C>T donor cells that were naïve or base edited. The variants from naïve and treated samples were compared for each donor. DETAILED DESCRIPTION Genome editing can be applied to restore endogenous gp91phoxexpression at physiological levels (11). To correct the CYBB c.676C>T mutation on the X chromosome, we utilized a CRISPR-Cas9 nuclease to create a targeted DNA double- strand break (DSB) with subsequent repair via homology-directed repair (HDR) using a short 100-nt DNA donor. However, relatively inefficient HDR and adverse effects of Cas9-induced DSBs have been reported, including activation of the DNA damage response, unwanted insertion or deletion mutations at the on-target site, large chromosomal deletions, translocations, chromothrypsis, and others (12). Next- generation genome editing technologies including base editors (BEs) overcome some of these concerns by correcting genetic mutations without intentionally introducing DSBs(13) (14-16). BEs are comprised of a Cas9 nickase fused to a DNA deaminase domain, which initiates the installation of C-to-T or A-to-G genome edits (via CBEs or ABEs, respectively). A guide RNA (gRNA) is expressed in trans to direct the BE to a target site that is flanked by a protospacer-adjacent motif (PAM). The efficiency, specificity, and product purity of BEs have been engineered and improved, supporting their use in preclinical studies(17)(18). However, the deaminase domain of BEs can typically only act within a short ‘edit window’ of the target site, restricting edit efficiency and flexibility. The edit window is typically nucleotides 4-9 of the spacer region of the target (17), which means that base edits are possible at a fixed distance from PAM of the target site. Thus, PAM availability is one of the major determinants of which regions of the genome are accessible for base editing. Attorney Docket No.29539-0835WO1 / MGH 2024-374 The conventional PAM for SpCas9 is NGG, where ‘N’ is any nucleotide (19- 21). For base editing approaches to correct point mutations, there is often not an NGG PAM available in the correct position of the target site to appropriately place the edit window of the BE over the edit of interest. Thus, alternative SpCas9 ‘PAM variant’ enzymes or alternate Cas9 orthologs are required to enable more precise positioning of the deaminase domain. An SpCas9 enzyme named SpRY was engineered to have a broadened and highly permissive PAM, primarily editing sites with NRN PAMs but also tolerating NYN PAMs (where R is A or G; Y is C or T) (22). Experiments using SpRY nuclease revealed efficient editing in cell lines (22) (22), plants (23-25), zebrafish (26), and C. elegans (25). Furthermore, SpRY was successfully adapted as both a CBE and ABE, and recently was combined with a higher activity ABE8e adenine deaminase domain (27) to enable more efficient A-to-G edits (23, 26). Described herein are efficient and precise base editing approaches to correct several CYBB mutations in patient-derived HSPCs. The use of ABEs overcomes many of the challenges associated with conventional lentiviral gene therapy or when using CRISPR-Cas nucleases with HDR donor templates. ABE-mediated editing results in improved cell viability, engraftment, and persistence of edited cells. The near PAMless base editor SpRY permits broader access to target sites encoding alternate PAMs, where high-fidelity mutations can improve safety by minimizing unintended edits. Transient delivery of fidelity-enhanced ABE-SpRY mRNA and gRNA via ex vivo electroporation into HSPCs resulted in durable on-target editing with minimal unwanted cis bystander or genome-wide off-target edits. The comprehensive pre- clinical package of efficiency and safety presented here, combined with the flexibility of high fidelity SpRY-based ABEs, should permit the extensibility of this blueprint to many other IEI-causative mutations for the development of prospective genetic treatments. BEs offer a potentially less genotoxic approach to install single nucleotide changes without intentionally generating DSBs or the requirement for an exogenous donor template. Many initial challenges of BEs have been overcome through iterative engineering (17, 18), including designs to improve on-target base editing efficiency(18, 27, 50) to minimize Cas-dependent and -independent off-target DNA and RNA editing, to reduce bystander editing, and to expand the applicability of BEs to a broader range of genomic sites (48, 51) (52, 53)(22). The present data Attorney Docket No.29539-0835WO1 / MGH 2024-374 demonstrated that the efficacy and precision of base editing in HSPCs offers advantages compared to LV-transduced cells, primarily by avoiding risks of LV integration near oncogenes. The continued development of methods to carefully monitor potential adverse events following infusion of BE-treated HSPCs, including how to measure biological consequences of off-target edits, will be necessary to ensure the safety of this new class of gene editing therapies. Transient delivery of ABE8e-SpRY-HiFi via mRNA into HSPCs ex vivo largely mitigated opportunities for off-target editing, minimizing most off-target editing events observed with ABE8e-SpRY. Via deep WES or via targeted sequencing of CHANGE-seq nominated sites, with ABE8e-SpRY-HiFi we observed no evidence of off-target editing in coding regions of the genome and only low levels of off-target editing at a small number of off-targets. Although the ABE8e-SpRY-HF1 construct had slightly attenuated on-target activity compared to ABE8e-SpRY-HiFi, it is possible that it could have further minimized or completely mitigated off-target editing. Of note, the WES reliably detected both the on-target edit and a bystander edit 2 basepairs upstream of the target / mutation. A caveat to interpreting this data is that the limited sampling of cells in a bulk highly polyclonal population (detection sensitivity of 0.4) may not be adequate to detect random, cell-specific mutations. Attempts to address this limitation have performed WES at a clonal level on large numbers of single cell colonies (54) or in clonally shrunken populations (55). Tens of thousands of mutation variants were reported following base editing; however, since similar numbers of substitutions were observed in mock samples, it is difficult to attribute the variants to the base editing process. The WES analysis herein demonstrated multiple potential sources for false positive variant calls including germline mutations that fail to be filtered out, which can be addressed using a more stringent bioinformatics pipeline. Although clonally shrunken populations of cells from limited colonies offer greater sensitivity to detect variants, such analyses remain susceptible to overcalling from failure to filter out germline mutations. At the RNA level, transcriptome analysis detected no off-targets other than an edit of a splice site in CYBB intron 6 that resulted in exon-skipping and aberrant exon 5-exon 8 junction at a low (~1%) efficiency. The absence of detectable transcriptome OTs in this study contrasts with previously results (48), likely attributable to transient Attorney Docket No.29539-0835WO1 / MGH 2024-374 mRNA delivery compared to plasmid delivery followed by sorting for highly- transfected cells with much higher levels of BE expression. Described herein are compositions and methods for highly efficient and specific base-editing to correct two distinct X-CGD mutations. Cas9 Base Editing Agents The present methods include the use of base editing agents to correct the CYBB mutations in HPSCs, preferably patient-derived HPSCs. The methods can include the use of expression vectors for in vitro or in vivo transfection and expression of a Cas9 base editor and suitable guide RNAs targeting CYBB. Alternatively, or in addition, the methods can include the in vitro or in vivo use of purified Cas9 base editor proteins complexed with suitable guide RNAs targeting CYBB. Nucleic Acids Encoding a CRISPR Gene Editing Complex The present methods can include the delivery of nucleic acids encoding a CRISPR gene editing complex. The gene editing complex includes a Cas9 base editing enzyme and one or more guide RNAs directing the editing enzyme to target CYBB. Guide RNAs directing the editing enzyme to CYBB The gene editing complex also includes guide RNAs directing the editing enzyme to a mutation in CYBB, i.e., comprising a sequence that is complementary to the sequence of a nucleic acid encoding a mutant CYBB, and that include a PAM sequence that is targetable by the co-administered Cas9 editing enzyme. Exemplary sequences for human CYBB can be found in GenBank, e.g., at NM_000397.4 (RefSeq nucleic acid), NP_000388.2 (RefSeq protein), and NG_009065.1, range 5001-38445 (RefSeqGene genomic sequence). Exemplary sequences of gRNAs targeting the mutations are provided below, e.g., in Table 1. CYBB is also known as cytochrome b-245 beta chain. The sequences provided herein are exemplary, and additional sgRNAs targeting the sequences identified herein could be designed and utilized. Cas9 base editing enzymes The methods include the delivery of Cas9 base editing enzymes to the cancer cells. The base editing enzymes can include the wild type SpCas9, or a SpRY variant, Attorney Docket No.29539-0835WO1 / MGH 2024-374 as well as variants thereof that are at least 80%, 85%, 90%, 95%, 99% or 100% identical thereto that retain at least one function of the parent case, e.g., the ability to complex with a gRNA, bind to target DNA specified by the gRNA, and alter the sequence of the target DNA. Preferably, to correct the c.676C>T mutation, the SpRY, SpRY HiFi, or SpRY HF1 Cas9 ABE is used, i.e., an SpCas9 variant capable of targeting NRN>NYN PAMs. SpRY BEs comprise mutations D10A / A61R / L1111R / D1135L / S1136W / G1218K / E1219Q / N1317R / A1322R / R1333P / R1335Q / T1337R); SpRY HiFi BE includes mutations D10A / A61R / R691A / L1111R / D1135L / S1136W / G1218K / E1219Q / N1317R / A1322R / R1333P / R1335Q / T1337R; the SpRY-HF1 includes mutations D10A / A61R / N497A / R661A / Q695A / Q926A / L1111R / D1135L / S1136W / G1218K / E12 19Q / N1317R / A1322R / R1333P / R1335Q / T1337R. Preferably to correct the c.1075G>A mutation, the SpG (comprising mutations D10A / D1135L / S1136W / G1218K / E1219Q / R1335Q / T1337R), WT SpCas9, or SpRY ABE is used. See, e.g., Walton et al., Science.2020 Apr 17;368(6488):290-296 and WO 2021151085. Other Cas9s can also be used for correcting either mutation, including SpG, SpCas9-NRTH, SpCas9-NRCH, and SpCas9-NRRH. Suitable target sequences for use with those Cas9s can readily be determined using known methods. The sequences of the Cas9s are known in the art; see, e.g., Walton et al., Science.2020 Apr 17;368(6488):290-296; Kleinstiver et al., Nature.2015 Jul 23; 523(7561): 481–485; WO 2021151085; WO 2016 / 141224; US 9,512,446; US-2014- 0295557; WO 2014 / 204578; and WO 2014 / 144761. The methods can also include the use of the other previously described variants of the SpCas9 platform (e.g., truncated sgRNAs (Tsai et al., Nat Biotechnol 33, 187-197 (2015); Fu et al., Nat Biotechnol 32, 279-284 (2014)), and nickase mutations (Mali et al., Nat Biotechnol 31, 833-838 (2013); Ran et al., Cell 154, 1380-1389 (2013)). Exemplary sequences are provided below. Base editors are known in the art and include cytosine base editors (CBE) and adenine base editors (ABE) that allow for the targeted deamination of cytosines and adenines, respectively, that are exposed on ssDNA by RNA-guided CRISPR-Cas proteins. Cytosine base editors (CBEs), such as BE3 or BE4max, catalyze the conversion of target C•G base pairs to T•A, while adenine base editors (ABEs), such Attorney Docket No.29539-0835WO1 / MGH 2024-374 as ABE7.10, ABEmax, or ABE8, convert target A•T base pairs to G•C. The Cas9 base editors (e.g., adenine base editors) used herein are fusion proteins comprising an adenosine deaminase domain and a Cas9 DNA binding domain. A number of variants of adenine base editors have been described, including ABEs 0.1, 0.2, 1.1, 1.2, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 2.10, 2.11, 2.12, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 4.1, 4.2, 4.3, 5.1, 5.2, 5.3, 5.4, 5.5, 5.6, 5.7, 5.8, 5.9, 5.10, 5.11, 5.12, 5.13, 5.14, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 7.1, 7.2, 7.3, 7.4, 7.5, 7.6, 7.7, 7.8, 7.9, 7.10, 8, 8e, 9, 9e, 8r ABEmax, e.g., as described in Gaudelli et al., Nature.2017 Nov 23; 551(7681): 464– 471; Koblan et al., Nat Biotechnol.2018 Oct;36(9):843-846); Tu et al., Mol Ther. 2022 Sep 7;30(9):2933-2941, Richter et al., Nat Biotechnol.2020 Jul;38(7):883-891; Chen et al., Nat Chem Biol.2023 Jan;19(1):101-110. Base editing with canonical base editors requires the presence of a PAM located approximately 15±2 base pairs from the target nucleotide(s). See, e.g., Komor, A.C. et al., Improved base excision repair inhibition and bacteriophage Mu Gam protein yields C:G-to-T:A base editors with higher efficiency and product purity, Sci Adv 3 (2017); Rees, H.A. et al., Improving the DNA specificity and applicability of base editing through protein engineering and protein delivery, Nat. Commun.8, 15790 (2017); US2018 / 0073012, US2017 / 0121693, WO2017 / 070633, US2015 / 0166980, U.S. Patent No.9,840,699; and U.S. Patent No.10,077,453. Split ABEs can also be used. Preferably, the CRISPR editing complex is specific, i.e., induces genomic alterations preferentially at the target site, and does not induce alterations at other sites, or only rarely induces alterations at other sites. Expression Constructs Expression constructs encoding one or both of guide RNAs and / or Cas9 base editing enzymes can be any composition capable of effectively delivering the component gene to cells in vivo. Approaches include insertion of the gene in viral vectors, including recombinant retroviruses, adenovirus, adeno-associated virus, lentivirus, and herpes simplex virus-1, or recombinant bacterial or eukaryotic plasmids. Viral vectors transfect cells directly; plasmid DNA can be delivered naked or with the help of, for example, cationic liposomes (lipofectamine) or derivatized (e.g., antibody conjugated), polylysine conjugates, gramacidin S, artificial viral Attorney Docket No.29539-0835WO1 / MGH 2024-374 envelopes or other such intracellular carriers, as well as direct injection of the gene construct or CaPO4precipitation carried out in vivo. A preferred approach for in vivo introduction of nucleic acid into a cell is by use of a viral vector containing nucleic acid, e.g., a cDNA. Infection of cells with a viral vector has the advantage that a large proportion of the targeted cells can receive the nucleic acid. Additionally, molecules encoded within the viral vector, e.g., by a cDNA contained in the viral vector, are expressed efficiently in cells that have taken up viral vector nucleic acid. In some embodiments, nucleic acids encoding a CRISPR base editing complex (e.g., a Cas9 BE or gRNA) are entrapped in liposomes bearing positive charges on their surface (e.g., lipofectins). These delivery vehicles can also be used to deliver Cas9 BE protein / gRNA complexes (e.g., RNPs). In clinical settings, the gene delivery systems for the nucleic acids encoding a CRISPR base editing complex can be introduced into a subject by any of a number of methods, each of which is familiar in the art. For instance, a pharmaceutical preparation of the gene delivery system can be introduced systemically, e.g., by intravenous injection, and specific transduction of the protein in the target cells will occur predominantly from specificity of transfection, provided by the gene delivery vehicle, cell-type or tissue-type expression due to the transcriptional regulatory sequences controlling expression of the gene, or a combination thereof. In other embodiments, initial delivery of the nucleic acids encoding a CRISPR base editing complex is more limited, with introduction into the subject being quite localized. In some embodiments, a controlled-release hydrogel comprising the nucleic acids encoding a CRISPR gene editing complex is administered to provide a steady dose of the nucleic acids encoding a CRISPR gene editing complex over time. A pharmaceutical preparation of the nucleic acids encoding a CRISPR gene editing complex can consist essentially of the gene delivery system (e.g., nucleic acids such as viral vector(s)) in an acceptable diluent, or can comprise a slow release matrix in which the gene delivery vehicle is embedded. Alternatively, where the complete gene delivery system can be produced intact from recombinant cells, e.g., retroviral vectors, the pharmaceutical preparation can comprise one or more cells, which produce the gene delivery system. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Recombinant Protein-RNA Complexes The Cas9 BE can be delivered as a purified protein (e.g., a recombinantly produced purified protein, prefolded and optionally complexed with the sgRNA, in a ribonucleoprotein complex (RNP)). Purified Cas9 proteins can be produced using methods known in the art, e.g., expressed in prokaryotic or eukaryotic cells and purified using standard methodology. See, e.g., Liang et al., Journal of Biotechnology 208:44–53 (2015); Kim et al., Genome Res.2014 Jun; 24(6): 1012–1019. Efficiency of protein delivery can be enhanced, e.g., using electroporation (see, e.g., Wang et al., Journal of Genetics and Genomics 43(5):319–327 (2016)); cationic or lipophilic carriers (see, e.g., Yu et al., Biotechnol Lett.2016; 38: 919–929; Zuris et al., Nat Biotechnol.33(1):73-80 (2015)); or even lentiviral packaging particles (see, e.g., Choi et al., Gene Therapy 23, 627-633 (2016)). Pharmaceutical Compositions The methods described herein can include the administration of pharmaceutical compositions and formulations comprising nucleic acids encoding base editors and gRNAs or recombinant base editor protein / gRNA RNP complexes targeting CYBB mutations as described herein, or cells that have been edited ex vivo to correct a CYBB mutation using a method described herein. In some embodiments, the compositions are formulated with a pharmaceutically acceptable carrier, such as a lipid nanoparticle (LNP), liposome, exosome, virus-like particle, or a variant thereof (see, e.g., Zhou and Yao, Mol Biomed 4, 10 (2023). doi.org / 10.1186 / s43556-023-00115-5; Banskota et al., Cell. 2022 Jan 20; 185(2): 250–265.e16; US11020470; Raguram et al., Cell. Volume 185, Issue 15, 21 July 2022, Pages 2806-2827; WO2020252455; WO2022020800; WO2024108001; WO2024107959 and WO2024107983). The LNPs can comprise various lipids, including dioleoylphosphatidylethanolamine (DOPE); 1,2-distearoyl- sn-glycero-3-phosphocholine (DSPC); 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC); 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC); 1,2-dipalmitoyl-sn- glycero-3-phosphocholine (DPPC); 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC); and 1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine (SOPC). See also Jung et al., Theranostics.2022; 12(17): 7509–7531. The pharmaceutical compositions and formulations can be administered parenterally, e.g., by intravenous administration. Attorney Docket No.29539-0835WO1 / MGH 2024-374 The pharmaceutical compositions can be formulated in any way and can be administered in a variety of unit dosage forms depending upon the condition or disease and the degree of illness, the general medical condition of each patient, the resulting preferred method of administration and the like. Details on techniques for formulation and administration of pharmaceuticals are well described in the scientific and patent literature, see, e.g., Remington: The Science and Practice of Pharmacy, 21st ed., 2005. Methods of Treatment The Cas9 BEs described herein targeting CYBB mutations can be administered for therapeutic treatments. In some embodiments, the compositions (e.g., nucleic acids or protein / gRNA complexes) are administered to a subject who has a disorder caused by an CYBB mutation, e.g., X linked Chronic granulomatous disease (X- CGD), in an amount sufficient to cure, alleviate or partially arrest the clinical manifestations of the disorder or its complications; this can be called a therapeutically effective amount. Tables 2 and 3 provides sets of exemplary combinations of BEs and gRNA. In some embodiments, a therapeutically effective amount is sufficient to produce a normal result on a dihydrorhodamine (DHR) test, which uses flow cytometry to measure the oxidation of dihydrorhodamine 123 to rhodamine 123 in phorbol myrisate acetate (PMA)-stimulated neutrophils, a marker for cellular NADPH oxidase activity (see, e.g., Vowells et al., J Pediatr.1996;128:104–7). The amount of pharmaceutical composition adequate to accomplish this is a therapeutically effective dose. The dosage schedule and amounts effective for this use, i.e., the dosing regimen, will depend upon a variety of factors, including the stage of the disease or condition, the severity of the disease or condition, the general state of the patient's health, the patient’s physical status, age and the like. In calculating the dosage regimen for a patient, the mode of administration also is taken into consideration. The dosage regimen also takes into consideration pharmacokinetics parameters well known in the art, i.e., the active agents’ rate of absorption, bioavailability, metabolism, clearance, and the like (see, e.g., Hidalgo-Aragones (1996) J. Steroid Biochem. Mol. Biol.58:611-617; Groning (1996) Pharmazie 51:337-341; Fotherby (1996) Contraception 54:59-69; Johnson (1995) J. Pharm. Sci. 84:1144-1146; Rohatagi (1995) Pharmazie 50:610-613; Brophy (1983) Eur. J. Clin. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Pharmacol.24:103-108; Remington: The Science and Practice of Pharmacy, 21st ed., 2005). The state of the art allows the clinician to determine the dosage regimen for each individual patient, active agent and disease or condition treated. Guidelines provided for similar compositions used as pharmaceuticals can be used as guidance to determine the dosage regiment, i.e., dose schedule and dosage levels, administered practicing the methods of the invention are correct and appropriate. Single or multiple administrations of formulations can be given depending on for example: the dosage and frequency as required and tolerated by the patient, the degree and amount of therapeutic effect generated after each administration (e.g., effect on tumor size or growth), and the like. The formulations should provide a sufficient quantity of active agent to effectively treat, prevent or ameliorate conditions, diseases or symptoms. In alternative embodiments, pharmaceutical formulations for oral administration are in a daily amount of between about 1 to 100 or more mg per kilogram of body weight per day. Lower dosages can be used, in contrast to administration orally, into the blood stream, into a body cavity or into a lumen of an organ. Substantially higher dosages can be used in topical or oral administration or administering by powders, spray or inhalation. Actual methods for preparing parenterally or non-parenterally administrable formulations will be known or apparent to those skilled in the art and are described in more detail in such publications as Remington: The Science and Practice of Pharmacy, 21st ed., 2005. The methods and compositions described herein can also be used for cell therapy. For example, stem cells from the patient are collected, corrected, and re- infused as a one-time intravenous infusion, a process referred to as cell therapy. The patient can receive chemotherapy using busulfan (which induces severe and prolonged myelosuppression) prior to cell infusion. Corrected stem cells engraft and grow into functional immune cells. Correction of at least about 10% of the myeloid cells can protect them from disease-specific infections. Exemplary sequences In some embodiments, the sequence of a protein or nucleic acid used in a composition or method described herein is at least 80%, 85%, 90%, 95%, 97%, 98%, or 99% identical to a sequence set forth herein. To determine the percent identity of two amino acid sequences, or of two nucleic acid sequences, the sequences are Attorney Docket No.29539-0835WO1 / MGH 2024-374 aligned for optimal comparison purposes (e.g., gaps can be introduced in one or both of a first and a second amino acid or nucleic acid sequence for optimal alignment and non-homologous sequences can be disregarded for comparison purposes). In a preferred embodiment, the length of a reference sequence aligned for comparison purposes is at least 80% of the length of the reference sequence, and in some embodiments is at least 90% or 100%. The amino acid residues or nucleotides at corresponding amino acid positions or nucleotide positions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, then the molecules are identical at that position (as used herein amino acid or nucleic acid “identity” is equivalent to amino acid or nucleic acid “homology”). The percent identity between the two sequences is a function of the number of identical positions shared by the sequences, taking into account the number of gaps, and the length of each gap, which need to be introduced for optimal alignment of the two sequences. The comparison of sequences and determination of percent identity between two sequences can be accomplished using a mathematical algorithm. For example, the percent identity between two amino acid sequences can be determined using the Needleman and Wunsch ((1970) J. Mol. Biol.48:444-453 ) algorithm which has been incorporated into the GAP program in the GCG software package (available on the world wide web at gcg.com), using the default parameters, e.g., a Blossum 62 scoring matrix with a gap penalty of 12, a gap extend penalty of 4, and a frameshift gap penalty of 5. In some embodiments, the sequence of a protein or nucleic acid used in a composition or method described herein has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions or deletions as compared to a sequence set forth herein. In some embodiments, the substitutions are conservative substitutions. EXAMPLES The invention is further described in the following examples, which do not limit the scope of the invention described in the claims. Materials and Methods The following materials and methods were used in the Examples below. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Study Design We hypothesized that the adenine base editor (ABE) ABE8e-SpRY-ABE8e can more efficiently correct mutations that cause X-linked chronic granulomatous disease (X-CGD) than previous CIRSPR nuclease and donor template approaches. Our objective was to investigate the efficiency, safety, and applicability of SpRY-BEs to correct two A>G mutations that cause X-CGD for eventual translation to the clinical setting. Blood samples were obtained with ethical approval (NIAID protocols 05-I-0213 and 94-I-0073), and informed consent from all patients was received in accordance with the Declaration of Helsinki. All samples had at least three biological replicates unless stated otherwise due to limited patient material. The novelty and investigational nature of this research did not allow predefined power calculation for sample size used. Due to the highly specific nature of the experiments, the operator conducting these experiments was not blinded to the samples’ conditions and samples were not randomized. Plasmids Plasmids used in this study are described in Table 1, with certain plasmids available on Addgene (addgene.org / Benjamin_Kleinstiver / ). SpCas9 gRNAs were cloned into pUC19-U6-BsmBI_cassette-SpCas9gRNA (20) (BPK1520; Addgene ID 65777); gRNA target sites and primers to amplify the genomic CYBB region are described in Table 4. Cell lines and cell culture Human HEK 293T cells (American Type Culture Collection; ATCC) were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated FBS (HI-FBS) and 1% penicillin-streptomycin. Samples of supernatant media from cell culture experiments were analyzed monthly for the presence of mycoplasma using MycoAlert PLUS (Lonza). Generation of HEK 293T cell lines To generate HEK 293T cell lines bearing a CYBB c676C>T mutation, transfections were performed 20 hours following seeding of 2x104HEK 293T cells per well in 96-well plates. Editing was first assessed in bulk populations of cells. Transfections contained 70 ng of CBE expression plasmids (Table 1) and 30 ng of a gRNA expression plasmid (Table 2) mixed with 0.72 µL of TransIT-X2 (Mirus) in a total volume of 15 µL Opti-MEM (Thermo Fisher Scientific), incubated for 15 Attorney Docket No.29539-0835WO1 / MGH 2024-374 minutes at room temperature, and distributed across the seeded HEK 293T cells. Cells were grown for approximately 72 hours prior to extracting genomic DNA (gDNA) by discarding the media, resuspending the cells in 100 µL of quick lysis buffer (20 mM Hepes pH 7.5, 100 mM KCl, 5 mM MgCl2, 5% glycerol, 25 mM DTT, 0.1% Triton X-100, and 60 ng / µL Proteinase K (New England Biolabs; NEB)), heating the lysate for 6 minutes at 65 ^C, heating at 98 ^C for 2 minutes, and then storing at -20 ^C. Editing efficiency in bulk transfected cells was assessed by next-generation sequencing (NGS) essentially as previously detailed (22) using PCR round 1 and round 2 primers (Table 4). Next-generation sequencing was performed to a depth of ~1,000+ reads per sample and base editing outcomes were analyzed using CRISPResso2(56). Following assessment in bulk populations of cells, to then create clonal cell lines, HEK 293T cells were seeded and transfected with 70 ng of plasmid expressing BE4max-SpRY (RTW5133) and 30 ng of the C6 gRNA expression plasmid (MNW752). Transfected cells were grown for approximately 72 hours prior to dilution plating into 96-well plates, which were then grown for ~2 weeks until confluent. Cells were transferred into 48-well plates with some cell mass reserved to extract genomic DNA (gDNA) for genotyping via PCR and Sanger sequencing or NGS (using CRISPResso2) to verify heterozygous or homozygous CYBB c676C>T mutation. CYBB c676C>T correction in HEK 293T cells To correct the CYBB c676C>T mutation in our heterozygous HEK 293T cell line, cells were seeded and transfected, genomic DNA extracted, and editing assessed as described above. Transfections contained 70 ng of ABE expression plasmid (Table 1) and 30 ng of a gRNA expression plasmid (Table 2). Human blood cells Human CD34+HSPCs were collected from healthy male donors or male X- CGD patients after written informed consent (NIAID protocols 05-I-0213 and 94-I- 0073). Base editing in CD34+HSPCs Cryopreserved CD34+HSPCs were thawed and cultured in StemSpan II supplemented with SCF, TPO and FLT3L (100ng / mL) and UM171 (35nmol / ml) for two days before base editing. Base editor mRNA and gRNA were delivered by Attorney Docket No.29539-0835WO1 / MGH 2024-374 electroporation (HPSC34-3 protocol; MaxCyte). HSPCs were cultured for an additional two days before harvest, or in some samples, a second electroporation with base editing reagents for a 2x-BE. For optimization studies, RNAse inhibitor was added to EP mix at concentrations as indicated, and at 1.6x103U / ml for all other studies. Glycerol was used at 2% final concentration. Molecular analysis of targeted editing in HSPCs Genomic DNA was extracted by DNeasy Blood & Tissue kit (QIAGEN) from HSPCs at 2-5 days post-EP for analysis of editing efficiency by high throughput sequencing (Illumina). Identification of candidate off-targets sites was performed in vitro on male X-CGD CD34+cells using CHANGE-seq for large-scale genome-wide profiling of off-targets (57). Targeted deep sequencing was performed using primers described in Table 4. In vitro myeloid differentiation and flow cytometric analysis Myeloid differentiation of CD34+HSPCs by culture in IMDM + 20% FSC for 10 -14 days before evaluation for gp91phoxexpression and NADPH oxidase function that are physiologically expressed only in mature phagocytes as previously described (58). (59). Expression of yH2AX (pS139) was performed using the Apoptosis, DNA Damage and Cell Proliferation Kit (BD Biosciences) according to manufacturer’s instructions. Transplantation of CD34+HSPCs into NSG mice CD34+HSPCs (CGD or HD naïve, base-edited) were transplanted into six- to eight-week old immunodeficient NOD.Cg-PrkdcscidIl2rgtm1Wjl / SzJ (The Jackson Laboratory, JAX5557) (NIAID Institutional Committee under animal use protocol LCIM-1E). Analysis was performed at 8, 12 and 18-26 weeks post-transplant on peripheral blood and bone marrow (18-26 weeks). Complete blood counts were monitored following transplant, and animal welfare formally documented to assess tolerability of transplants with BE-HSPCs. Experimenters were not blinded to the sample conditions injected into mice, and mice were randomized to treatment groups. CHANGE-seq CHANGE-seq was performed as previously described (44). Briefly, genomic DNA from human primary HSPCs was isolated using Gentra Puregene Kit (Qiagen) Attorney Docket No.29539-0835WO1 / MGH 2024-374 according to manufacturer's instructions. Purified genomic DNA was tagmented with a custom Tn5-transposome to an average length of 400 bp, followed by gap repair with Kapa HiFi HotStart Uracil+ DNA Polymerase (KAPA Biosystems) and Taq DNA ligase (NEB). Gap-repaired tagmented DNA was treated with USER enzyme (NEB) and T4 polynucleotide kinase (NEB). Intramolecular circularization of the DNA was performed with T4 DNA ligase (NEB) and residual linear DNA was degraded by a cocktail of exonucleases containing Plasmid-Safe ATP-dependent DNase (Lucigen), Lambda exonuclease (NEB) and Exonuclease I (NEB). In vitro cleavage reactions were performed with 125 ng of exonuclease-treated circularized DNA, 90 nM of SpRY and SpRY-HF1 protein (purified as previously described)(62), NEB buffer 3.1 (NEB) and 270 nM of sgRNA, in a 50 μL volume. Cleaved products were A-tailed, ligated with a hairpin adaptor (NEB), treated with USER enzyme (NEB) and amplified by PCR with barcoded universal primers NEBNext Multiplex Oligos for Illumina (NEB), using Kapa HiFi Polymerase (KAPA Biosystems). Libraries were quantified by qPCR (KAPA Biosystems) and sequenced with 151 bp paired-end reads on an Illumina NextSeq instrument. CHANGE-seq data analyses were performed using open-source CHANGE-seq analysis software (github.com / tsailabSJ / changeseq). Off-target site sequencing via rhAmpSeq For off-target validation studies, CYBB c.676C>T CGD patient HSPCs were edited via electroporation (as described above) using ABE8e-SpRY or ABE8e-SpRY- HiFi mRNA and gRNA A5 or remained unedited (naïve). gDNA was extracted (DNeasy Blood & Tissue Kit; Qiagen). The on-target site and top 50 CHANGE-seq nominated off-target sites were analyzed via rhAmpSeq according to the manufacturer’s instructions (Integrated DNA Technologies), prior to sequencing on an Illumina sequencing platform. CRISPResso2 was used to quantify base editing efficiencies. Using the amplicon primer coordinates from the rhAmpSeq pool, the amplicon sequences for each of the on- and off-target sites were extracted using the UCSC genome browser (genome.ucsc.edu / ). For each site, the chromosome position of the amplicon was inputted as a custom track and the amplicon sequence was obtained using the table browser function, selecting the sequence annotation output format with hg38 assembly as the chosen dataset. These amplicon sequences were used for analysis of the rhAmpSeq data from a 300c NextSeq sequencing run, as input Attorney Docket No.29539-0835WO1 / MGH 2024-374 into CRISPRessoPooled (v.2.0.30) to calculate the efficiency of base editing using the following commands: --min_reads_to_use_region 100 -w 20 --cleavage_offset -10 --base_editor_output --conversion_nuc_from A --conversion_nuc_to G. For both on- and off-target sites, the quantification window nucleotide percentage table outputs were merged across treated and untreated control samples to streamline analysis. The A-to-G and C-to-T / G / A base editing efficiencies at adenines and cytosines in positions 1 through 12 of the target site spacer (counting from the PAM distal end of the spacer) were determined and graphed using GraphPad Prism (v.10.2.3). WES analysis We combined unique molecular barcodes (MBCs) with ultra-deep Illumina sequencing to achieve high sensitivity for rare variant detection. The first step in WES library preparation involves fragmentation of genomic DNA. We used two methods, enzyme fragmentation (EF) and Covaris shearing (CS) library protocol to make libraries in duplicates for each sample to avoid low frequency errors that could be introduced by specific fragmentation method. For WES, Agilent SureSelect HS2 human exon V7 kit with molecular barcode was used. BE and non-edited (naïve control) samples from two patients were processed in duplicates, for a total of 8 samples sequenced on Illumina NextSeq2000 P3-300 kit (2 x 150 PE reads, 2.4 billion reads). For the BE samples, the coverage depth was about 500x for each library, with 1000x combined for each sample prepared by enzyme fragmentation(EF) and Covaris shearing (CS) library protocol. This allows us to detect potential low frequency mutations to less than 1% (3 supported reads / 500 reads). For bioinformatics analysis, two pipelines were used. First one was similar to Fiumara et al(55), which variant calls for each sample were made against human reference genome hg38, and then variants in the naïve samples are subtracted from BE samples. For this pipeline, Agilent SureCall software was used, which was designed for low frequency variant calling with the Agilent SureSelect HS2 with molecular barcode kit. We first used Agilent command line tool AGeNT2 to extract molecular barcode for each read. The MBC is the first 3-5bp on both ends of the read (ends with CT or GT due to T / A ligation). The combined trimmed MBC is 3bp-3bp RX tag and gives 46different barcodes. After extraction and trimming of MBC on the end of reads, the resulted fastq files were used to map to the human genome hg38 Attorney Docket No.29539-0835WO1 / MGH 2024-374 using SureCall. The settings are: keep all duplicates (we can confirm with MBC later). Alignment was done with BWA-MEM. Only Unique best hits were reported. Variant score threshold was set at 0.001 and variant will be supported by two or more reads. The output is BAM files and VCF files with variant calling. This pipeline generated too much “treatment specific” mutation calls, most of which are germline variants failed to be filtered based on that the majority overlap with known SNPs. A second pipeline was a modified version of the OpenOmics genome-seek pipeline (github.com / OpenOmics / genome-seek) to perform variant calling jointly in pairs of control and edited samples. In brief, the Agilent AGeNT v3.0.5 software (agilent.com / en / product / next-generation-sequencing / hybridization-based-next- generation-sequencing-ngs / ngs-software / agent-232879) was used to trim reads and perform MBC-based deduplication, reads were then mapped to the GRCh38.p14 reference genome with bwa-mem2 (Vasimuddin et al.2019), and deduplicated BAM files underwent Indel realignment with GATK v3.8 and quality recalibration using GATK4 v4.4.0.0 (Van de Auwera and O’Connor 2020). Recalibrated BAM files were then used to perform paired variant calling with the non-edited naïve samples serving as the control using four variant callers: Octopus v0.7.4 (63)(Cooke et al.2021), Strelka2 (64), Muse (65) and MuTect2 (Benjamin et al.2019 BioRx). Variants were initially filtered to include only those called by at least 2 variant callers, variants with >=3 reads supporting the mutant allele, and variants rare in the general population (<0.001 in Gnomad v3 and 1000 Genomes). All somatic variants were then annotated The second WES analysis pipeline identified only 15 somatic mutations in all treated samples. Manual inspection of the 15 mutations using IGV revealed that 9 of the 15 mutations were based on sequencing reads in only one direction and eight of these were found only in results from one of the two library preparation protocols. False positive variant filtering based on strand biased support is an established practice demonstrated to be highly effective at removing artifacts (66). This also suggests that these type of low frequency mutations are largely specific for the CS library preparation process, most likely due to the DNA damage to one of the strand during ultrasonic shearing. Bona fide mutations should be supported by both strands and in both CS and EF libraries since these are simply replicates of the same DNA sample. Of the remaining 6 of the 15 mutations called, 3 were found in naïve sample with a single read support, while 2 were known SNPs at very low frequency. There was only Attorney Docket No.29539-0835WO1 / MGH 2024-374 one mutation at chr12:57727421 in the AGAP2 gene that was absent in naïve but present in all 4 treated libraries (both CS and EF). The frequency of this variant, however, was low (3 / 751, 12 / 760, 3 / 722, and 2 / 755) in all of the treated samples. This could be a bona fide BE-induced variant, or potentially attributable to sequencing error / noise due to lower coverage (100-200x) in the naïve samples. Karyotyping and Fluorescent In-Situ Hybridisation (FISH) Cells were prepared for chromosomal analysis and stained for fluorescence in- situ hybridization (FISH) as previously described(60), and images were acquired using an epi-fluorescence microscope (Imager Z2, Zeiss) and HiFish acquisition software version 8.2 (Genasis, Applied Spectral Imaging,Inc., CA).(61) (61) Optical genome mapping Optical genome mapping was performed by Bionano Genomics and was applied to base edited CGD HSPC and a naïve CGD control. To evaluate the sensitivity of the OGM technology, DNA from a Kasumi-1 leukemia cell line with known translocations including t(8:21)AML1-ETO were spiked in to the test (patient DNA). We confirmed detection of the t(8:21) translocation below 5% VAF at 1.5 Tb coverage and at 5 Tb, the increased coverage detected the known translocation can be detected at 1% VAF. Our BE samples were tested at 5 Tb coverage. COMET assay DNA damage was detected by alkaline single-cell gel electrophoresis (SCGE). Briefly, 10,000 cells in 1% agarose were spread onto microscopic slides pre-coated with 1% agarose. The cells were lysed for 1 hour at 4 °C in an alkaline buffer (2.5 M NaCl, 0.1 M Na2EDTA, 10 mM Tris–HCl and 1% Triton X-100, pH = 10). Electrophoresis at ph≥13, 4 °C for 30 minutes at 1V / cm. The slide was immersed in distilled water for 5 minutes, washed with 70% ethanol, and dried. Gel Green Nucleic Acid Stain was added to dried agarose for 15 minutes, then analyzed for visualization and quantification of DNA breaks by fluorescence microscopy. RNA sequencing RNA-seq analysis was performed by first mapping reads to human genome hg38 using STAR, Picards Tools for quality control and GATK4.0 called out variants (254,086 in naïve sample, 467,102 in edited sample)(“Picard Toolkit.” 2019. Broad Institute, GitHub Repository. (broadinstitute.github.io / picard / ). Baseline variants in naïve were filtered out, leaving about 90,000 A>G single nucleotide changes. When restricted to A>G changes with Attorney Docket No.29539-0835WO1 / MGH 2024-374 >100 read-coverage, 239 variants were called but none were found to be real editing events due to low coverage or frequency too low to reach threshold for variant frequency. To evaluate exon skipping, we collected all reads that spanned the exons 6, 7, and 8 junctions were collected since the bystander edit affects the splice acceptor immediately upstream of CYBB exon 7. The frequency of exon 6-7 (normal splicing) vs exon 6-8 or 5-8 (skipping splicing) estimated that only 1% of exon junctions were indicative of skipping splicing events ((exon5-exon8)+(exon6-exon8)) / (exon6-exon7). Statistical analysis Non-parametric Mann-Whitney test were used to compare frequencies of corrected alleles from targeted sequencing results, gp91phoxand DHR expression analyses. Error bars depicted in graphs represent standard deviations of ≥ 2 independent biological replicates. A p value of ≤ 0.05 was considered statistically significant. Annotations above individual bars represent the mean of the plotted values. Mean and SD were calculated for experiments with multiple replicates under the same conditions unless stated otherwise. All statistical analyses were performed using GraphPad Prism 9.4.1 for macOS (GraphPad Software, San Diego, CA). Table 1. Plasmids used in this study Addgene plasmid plasmid description ID nickname T 1 P T 2 / Attorney Docket No.29539-0835WO1 / MGH 2024-374 Addgene plasmid plasmid description ID nickname D1135N / D1180G / G1218S / E1219V / Q1221H / P1249S / E125 D 3 8 Table 2. gRNAs and target sites used in this study 4 nt spacer site description gRNA spacer sequence # PAM length (nt) Attorney Docket No.29539-0835WO1 / MGH 2024-374 4 nt spacer site description gRNA spacer sequence # PAM length (nt) CYBB 7 t Target site Attorney Docket No.29539-0835WO1 / MGH 2024-374 Target site protospacer (20 nt) - gRNAs used to create mutations 4 t ) Attorney Docket No.29539-0835WO1 / MGH 2024-374 Table 3. BE / gRNA combinations CYBB c.676C>T (p.R225X) top enzymes and gRNAs Base editor gRNA Table 4. Primer and Amplicon Sequences # P5 primer # P5 rimer full annealin 5 7 9 Attorney Docket No.29539-0835WO1 / MGH 2024-374 ACACTCTTTCCCTACACGACGC 60 GTACAGGGCCTA 61 CYBBexon7_create_c6 TCTTCCGATCTGTACAGGGCCT CATCAGAGCACT 76C>T-CBE-C6-SpCas9 ACATCAGAGCACTT T 3 5 7 9 1 3 5 # P7 primer # P7 primer full annealing 7 9 1 3 5 7 Attorney Docket No.29539-0835WO1 / MGH 2024-374 CYBBexon7_correct_c GACTGGAGTTCAGACGTGTGCT 88 89 676C>T-ABE-A5- CTTCCGATCTCCAGCAAACTGA CCAGCAAACTGA 1 3 5 7 site # description amplicon sequence 0 1 2 3 Attorney Docket No.29539-0835WO1 / MGH 2024-374 GTACAGGGCCTACATCAGAGCACTTAAAATATATGCAGAATCTTTTAA 104 CYBBexon7_ TAAAACAATTTAATTTCCTATTACTAAATGATCTGGACTTACATTTTT correct c6 CACCCAGACGAATTGTACGTGGGCAGACCGCAGAGAGTTTGGCTGTGC 5 6 7 8 Example 1. Modeling and correction of CYBB c.676C>T in HEK 293T cells To explore BE-mediated correction of IEI-causative mutations, we first created a cell-based model using CBEs to harbor a common c.676C>T mutation within CYBB exon 7 that is known to cause X-CGD (Figs.1a-1c). We utilized HEK 293T cells due to ease of transfection to simplify the evaluation of various BE constructs and gRNAs, leading to the creation of clonal cells bearing the specific CYBB c.676C>T mutation (Fig.1c and Figs.9a-9f). Next, to evaluate ABE-mediated corrective approaches (Fig.1d), we designed a series of gRNAs that tile the ABE edit window across the target CYBB c.676C>T mutation (denoted by target nucleotide positions of A3 through A9 numbered from the PAM proximal end of the individual spacers for each gRNA; in Fig.1e). Notably, none of these target sites encode an NGG PAM (Fig.1e), precluding the use of the conventional wild-type (WT) SpCas9-based ABEs. Instead, we utilized ABE8e-SpRY that should in principle tolerate each PAM encoded by the various target sites. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Transfections were performed in a heterozygous CYBB c.676C>T cell line using plasmids encoding ABE8e-SpRY and each gRNA (Fig.10) led to correction of the c.676C>T mutation with each gRNA (Fig.1f). With certain gRNAs, we observed low-level unwanted ‘bystander’ editing of nearby adenosine or cytosine bases (Figs. 1g, 1h, and Fig.11a), similar to as previously described for ABEs when the C of a TC motif falls in positions 5 or 6 of an ABE target site (30). A potential caveat of engineered SpCas9 enzymes with relaxed PAM tolerances is the enhanced likelihood of observing genome-wide off-target edits (due to an expanded genomic search space potentially leading to additional off-target encounters)(22,24,31). To minimize potential off-target edits, we combined SpRY with previously described high-fidelity mutations in SpCas9 that reduce off-target editing at mismatched sites (28, 29). When testing ABE8e-SpRY-HF1 and ABE8e- SpRY-HiFi constructs, with most gRNAs we observed similar levels of on-target editing compared to ABE8e-SpRY (Fig.1i). We also observed that both SpRY-HF1 and SpRY-HiFi constructs led to somewhat reduced bystander editing (Fig.11b). Next, since gRNA A7 harbors an NGT PAM, we also explored the use of other previously described engineered SpCas9 PAM variant enzymes that might have improved or similar genome-wide specificities compared to SpRY. We observed that SpG (22) and SpCas9-NRTH (32) offered similar or slightly higher levels of on-target editing compared to SpRY, though could also lead to increased bystander editing of nearby adenines (Fig.12), presumably due to their increased on-target efficiencies. Example 2. Optimization of base editing in human healthy donor CD34+hematopoietic stem cells To conserve limited numbers of patient CD34+HSPCs for pre-clinical scale- up validation studies, to optimized base editing in HSCPs from healthy human donors (HD) we developed a surrogate editing approach using a different gRNA. We envisioned evaluated editing in HD HSPCs using ABE8e-SpRY with the surrogate gRNA targeting an intronic splice acceptor site in CYBB intron 6 (named i6-2), located 3 nucleotides upstream of the CYBB exon 7 c.676C>T mutation (Fig.2a). This surrogate target site permits the use of the same base editor (ABE8e-SpRY) and targeting the splice site provides a functional readout via loss of gp91phox expression in base edited cells. This workflow should permit the evaluation of key parameters of Attorney Docket No.29539-0835WO1 / MGH 2024-374 HSPC culture and electroporation using HD HSPCs, where the optimized methods could later be translated into patient-derived cells. Peripheral blood mobilized CD34 HSPCs were electroporated (EP) with mRNA encoding ABE8e-SpRY and a synthetic i6(-2) gRNA, and we assessed A-to-G editing at CYBB intron 6 (-2). Cells were further cultured for 2 days following EP prior to harvest for analysis. We evaluated various HSPC culture and stimulation conditions prior to EP. Conventionally, CRISPR-Cas9 nucleases are delivered to HSPCs after 2 days of ex vivo culture with cytokines (SCF, TPO and FLT3L) to induce cell proliferation for optimal delivery of genome editing reagents and cell- cycle dependent HDR-conversion to desired genetic sequence (Fig.2b). Since BEs are not thought to rely on cell cycling (33), we wondered whether ex vivo cell culture could be shortened in duration in order to preserve cell stemness and engraftment capability. Interestingly, the efficiency of base editing as determined by next- generation sequencing (NGS) was superior after 2 days ex vivo culture compared to a single-day culture (p = 0.0303) (Fig.2c). These results suggest a potential importance of genomic accessibility for efficient base editing. Auxiliary reagents such as UM171 that promotes HSPC expansion (34) or the cytoprotectant glycerol previously shown to be beneficial for CRISPR-Cas9 nuclease-mediated editing of HSPCs (11) did not substantially alter base editing efficiency in HSPCs (Fig.2d). Successful base editing of the surrogate CYBB i6 (-2) site at the splice acceptor preceding exon 7 can result in the disruption of CYBB expression of gp91phox protein (Fig.2a). As the CYBB gene is primarily expressed in mature phagocytes, CD34+ HSPCs were myeloid-differentiated ex vivo before evaluating gp91phox expression by flow cytometry. The loss of gp91phox expression in base- edited samples indicated successful editing and disruption of the splice acceptor (Fig. 2e). Since the editing process appeared well tolerated, we evaluated the impact of a repeat EP of ABE8e-SpRY mRNA with gRNA i6(-2) two days following the first EP. The cells underwent further ex vivo culture for two days after the 2ndEP before harvest for cryopreservation and analysis. HSPCs base-edited twice showed a greater loss of gp91phoxexpression (67%) compared to 1x EP (42%) myeloid differentiated cells (Fig.2e). To determine if base edited HSPCs engraft and persist long term, ABE8e- SpRY and gRNA i6(-2) treated HSPCs (BE-HSPCs) from HDs (n=3 subjects) were Attorney Docket No.29539-0835WO1 / MGH 2024-374 transplanted into immunodeficient NSG adult (6-8 week old) mice (1.8 to 2.3 x106 / animal). Evaluation of mice peripheral blood for gp91phox+ myeloid cells revealed stable knock-downs in gp91phoxexpression following transplant (Fig.2f). Human CD45+engraftment rates in spleen and bone marrow showed no significant difference between 2xEP (20.8%, n=8) or 1xEP (24.7%, n=15) in bone marrow (Fig. 2g). Myeloid differentiated bone marrow cells reveal a slightly greater loss of gp91phoxexpression in CD15+myeloid cells after 2x EP (67.6%) than 1x EP (42.2%) relative to 95.1% in naïve HD cell, approximately a 60% knock-down in gp91phoxexpression (Fig.2e). NADPH oxidase release of reactive oxidative species (ROS) upon stimulation reduces dihydrorodamine to fluorescent rhodamine detectable by flow cytometric assay (DHR), which revealed a similar level of knock-down in function in 1x- and 2x-EP HSPCs (Fig.2h). Engrafted HSPCs differentiated into multi-lineage immune cells in normal distribution compared to HD HSPCs in bone marrow (Fig.2i) and spleen (Fig.2j). Example 3. Characterization of base edited X-CGD CD34+HSPCs After establishing our workflow in HD HSPCs, we then proceeded to evaluate base editing in X-CGD patient-derived CD34+HSPCs using ABE8e-SpRY mRNA and synthetic gRNAs A4-A7 targeting CYBB c.676C>T (Fig.1e). Assessment of on- target editing after 1x or 2x EP revealed a range of efficiencies with gRNAs A4 (mean: 42.3%) and A5 (mean: 55.9%) achieving the highest A-to-G conversion (Fig. 3a), consistent with our results from our HEK 293T experiments (Fig.1f). A prior study showed that two rounds of base-editing in human HSPCs yielded improved editing efficiency, albeit at the cost of initial reduction of cell viability (35). Confirmation of editing via NGS revealed a significant increase in editing with the 2x EP from 42.3% to 52.5% (Mann-Whitney test, p=0.0047) with c.676C>T gRNA A4, and an increase from 55.9% to 69.2% (Mann-Whitney test, p=0.1143) with c.676C>T gRNA A5 (Fig.3a). In previous CRISPR nuclease experiments, chemical modification of DNA donor molecules or synthetic gRNAs significantly improved editing efficiencies by reducing the rate of degradation of nucleic acid components(11, 36). Given our all-RNA approach, we added RNase inhibitor to our electroporation mixture to dampen RNA degradation by contaminating RNases, which led to modest improvements in editing efficiencies (Fig.3b). Attorney Docket No.29539-0835WO1 / MGH 2024-374 A primary driving motivation for targeted gene correction is the restoration of physiological expression of the target gene, compared to gene therapies that result in non-natural and suboptimal expression of transgenes via exogenous promoters and regulatory elements (2). Our assessment of myeloid differentiated CD34+ HSPCs revealed expression of gp91phoxprotein at similar mean fluorescent intensities as healthy donors, indicating the same amount of protein expression on a per cell basis as HD cells, consistent with restoration of physiological expression (Fig.3c). Correction of gp91phox expression also restored NADPH oxidase complex function for ROS production detected by a DHR flow cytometry assay (Fig.3d). In cells treated with ABE8e-SpRY mRNA and the c.676C>T A5 gRNA, we observed gp91phox+cells in myeloid-differentiated HSPCs in vitro were 51.7% (n=13) after 1x EP and 59.0% (n=5) after 2x EP (Fig 3e). Example 4. Extensibility of ABE-based correction to other CYBB mutations To explore the versatility of our ABE8e-based editing approach, we next extended our approach to target a second CYBB mutation c.1075 G>A in an X-CGD patient who does not have a matched donor for allogeneic stem cell transplant. For this mutation, there is an array of target sites bearing various PAMs (Fig.4a). Although three of the target sites harbor conventional NGG PAMs that place the target adenine within the canonical ABE8e edit window, we also explored the use of PAM variant enzymes with other gRNAs, due to the presence of nearby adenines where bystander editing may create variants of uncertain functional consequences. In experiments using CYBB c.1075G>A patient-derived HSPCs electroporated with ABE8e-SpG with c.1075G>A gRNA A5, ABE8e-WT with gRNAs A6-A8, or ABE8e-SpRY with gRNA A9, we observed comparably high on-target editing ranging from ~80 to >95% (Fig.4b). With ABE8e-SpRY-HF1 and c.1075G>A gRNA A5 we observed lower levels of editing (~48%; Fig.4b). We confirmed restoration of gp91phox expression and NADPH oxidase function by FACS, demonstrating gp91phoxexpression in BE-HSPCs with CYBB c.1075G>A mutation at similar levels compared to healthy donors (96%), and over 80% DHR+ cells consistent with restored NADPH oxidase function (Figs.4c, 4d). This restoration of protein expression and functional correction is attributed to high levels of precise correction, with only very low levels of bystander edits resulting in a non- Attorney Docket No.29539-0835WO1 / MGH 2024-374 synonymous ATC>GTC change resulting in I357V, and a non-synonymous GAC>GGC causing in D360G (Fig.4e). Using cell surface markers (CD34, CD38, CD90 and CD133), we compared relative frequencies of CD34 subsets following 1x- or 2x-EP, and noted a decrease in CD34+38- subset but maintained the most primitive CD34+CD38-CD90+CD133+subset proportions despite two-time EP in CYBBc.1075 mutation patient HSPCs (Fig.4f). A similar experiment using HD cells electroporated with mRNA encoding ABE8e-SpRY and gRNA i6(-2) showed similar maintenance of CD34 subsets after 1x (green) or 2x (blue) electroporation (Fig.4g). Overall, the higher levels of correction for CYBB c.1075G>A with ABE8e- WT with gRNA A8 (mean, 84.4%; n=7) compared to c.676C>T correction when using ABE8e-SpRY with gRNA A4 (mean, 49.5%; Mann-Whitney test, P = 0.0006; n=19) or gRNA A5 (mean, 51.7%; Mann-Whitney test, P = 0.0020; n=10) highlight the variability of editing between targets (possibly due to the greater binding affinity of wild-type SpCas9 ABEs compared to SpRY, gRNA folding, or other target- specific features), not withstanding that both editing approaches yielded sufficiently high levels of editing for robust phenotypic recovery and potential clinical translatability (Fig.4h) (24, 37, 38). Example 5. Analysis of base editing in CD34+HSPC subpopulations In vitro assessment Next, we sought to corroborate this observation of maintenance of editing in primitive HSPCs initially in experiments using HD cells. Quiescent primitive HSPCs account for long term engraftment following transplant and the ability to edit true HSPCs remains a critical objective to envsure persistence of genetic correction. To compare the ability to edit more quiescent CD34 subsets, we treated healthy donor cells treated with ABE8e-SpRY mRNA and the surrogate CYBB i6 (-2) gRNA. Electroporated cells were sorted them into CD34+subpopulations based on cell surface markers CD34, CD38 (39), CD45RA (40), CD90(41) and CD133 into CD34 subsets common myeloid progenitor (CMP CD34+CD38+CD45-), multi-potent progenitor (MPP CD34+38-CD45RA-CD90-), MLP (CD34+ CD38- CD45RA+), and HSC (CD34+CD38-CD45RA-CD90+CD133+)CD34+CD38+common lymphoid progenitor, CD34+38-CD45RA-CD90- multi-potential progenitor MPP), and CD34+CD38-CD45RA-CD90+CD133+HSC subpopulations (42). Analysis of editing via NGS revealed similar editing via the surrogate i6 (-2) gRNA across all sorted Attorney Docket No.29539-0835WO1 / MGH 2024-374 CD34+subpopulations, including the most primitive CD34+38-45RA-90+133+HSC subset after a single EP and was preserved even following the 2x EP protocol (Fig. 4g). This suggested that the editing process did not. substantially affect the distribution of CD34+subsets that is often associated with cell stemness and ability to engraft. In vivo xenotransplants To evaluate engraftment potential and long-term persistence of gene correction, human patient-derived CD34+HSPCs bearing the CYBB c.676C>T mutation base edited with ABE8e-SpRY mRNA and gRNA A5 were transplanted into immunodeficient NSG mice that were analyzed by peripheral blood draws approximately every 4 weeks until final harvest at 20-23 weeks (Fig.5a). Serial analysis of mice peripheral blood by flow cytometric analysis for gp91phox+cells revealed stable levels of engraftment (human CD45+cells ~10%) after an initial decline with ~60% of engrafted cells demonstrating restored gp91phox expression in mice transplanted with BE-treated patient HSPCs (Figs.5b-5d). Notably, we observed much higher levels of phenotypic correction in human CD45+ cells from mice transplanted with BE-treated cells compared with our previously reported Cas9 nuclease-based method using a single stranded DNA donor for HDR-mediated correction (Fig.5d) (11). Peripheral blood NADPH oxidase activity at 20-23 weeks in transplanted showed comparable DHR+ cell percentages from either 1x or 2x EP conditions that reached similar levels of NADPH oxidase function to HD control cells (~52% and ~46% for 1x and 2x EP, respectively, compared to ~60% for cells from mice transplanted with HD cells) (Fig.5e). At harvest (20-23 weeks), similar levels of engraftment of HSPCs were observed in 1x EP cells treated with ABE8e-SpRY and CYBB c.676C>T gRNA A5 compared to naïve patient HSPCs, suggesting that electroporation or base editing did not interfere with engraftment in bone marrow (mean of 27% (1x EP, n=14) versus 61% (2x EP, n=7), n=2 experiments) (Fig.5f). Somewhat higher engraftment observed with doubly electroporated HSPCs was unexpected, since extended ex vivo culture promotes HSPC differentiation and loss of engraftment potential. However, recent advances in small molecule engineering such as UM171 have improved HSC proliferation while maintaining stemness especially in cord blood that are in Phase I / II clinical trials (43). The cell expansion from the extra two days of ex vivo culture in Attorney Docket No.29539-0835WO1 / MGH 2024-374 presence of UM171 likely more than compensated for losses secondary to the additional manipulation and ex vivo culture. These observations warrant further investigation in future studies. Myeloid-differentiated bone marrow (BM) cells of mice transplanted with HSPCs treated with ABE8e-SpRY and CYBB c.676C>T gRNA A5 confirmed ~55% rates of gp91phox+ and NADPH oxidase functioning DHR+cells (Fig.5g). Of note, cells from mice originally treated via 1x and 2x EP expressed similar levels of gp91phox+, but the percentage of DHR+ cells was slightly higher from mice transplanted with 2x EP HSPCs (Fig.5h). Edited HSPCs retained ability to develop into immune cell lineages in peripheral blood (Fig 5h) and bone marrow (Fig.5i) which supports normal differentiation capacity of base-edited HSPCs. Similar phenotypic outcomes were observed for our second X-CGD correction approach of CYBB c.1075G>A with ABE8e-WT mRNA and gRNA A8, except at higher levels of mean correction with 54% to 62% gp91phoxexpression (after 1x or 2x EP, respectively, to 64% or 74% of HD levels; Fig.5j). Mice transplanted with CYBB c.1075G>A base edited HSPCs displayed gp91phoxexpression at 62% to 39% and 35.6% and 17.7% to 4.9% DHR+in myeloid-differentiated bone marrow cells (Fig. 5k). Notably, proof of principle correction for CYBB c.1075G>A was achieved more quickly using the same methodology as CYBB c.676C>T since the same base editing approach was utilized. Example 6. Safety considerations for ex vivo base editing We investigated potential consequences of base editing HSPCs, including unintended edits at off-target sites located throughout the genome, unwanted edits of nearby DNA bases within the target site (bystander edits), and potential cellular toxicities or genomic instability caused by the base editor or electroporation protocol. To reduce the likelihood of unwanted genome-wide edits, we tested the compatibility of ABE8e-SpRY-HF1 and ABE8e-SpRY-HiFi with gRNA A5 in CYBB c.676C>T HSPCs. Unlike in HEK 293T cells where we observed comparable performance between the SpRY and SpRY high-fidelity variants via plasmid-based transfection (Fig.1g), when electroporating a typical dose of ABE mRNA into HSPCs, we observed that the high-fidelity variants resulted in somewhat reduced on-target correction (Fig.6a). Interestingly, when electroporating a 2x dose of ABE8e-SpRY- HiFi mRNA, we then observed higher levels of on-target correction at levels more Attorney Docket No.29539-0835WO1 / MGH 2024-374 comparable to ABE8e-SpRY editing efficiency (Fig.6b). The higher dose of ABE8e- SpRY-HiFi mRNA (400 ug / mL EP) retained undetectable levels of bystander editing (Fig.6b). These results reveal that the high fidelity Cas9 enzymes are compatible with ABE8e-SpRY when delivered via mRNA into primary human HSPCs, though their use may necessitate an increased dose to achieve comparable levels of editing to ABE8e-SpRY. Of note, the levels of mutation repair remained stable in spleen and bone marrow following transplant into mice (Fig.6c). In experiments in CYBB c.1075G>A HSPCs, the use of a high fidelity ABE8e-SpCas9-HF1 mRNA along with the c.1075 gRNA A8 led to comparable levels of on-target correction compared with ABE8e-WT at an equivalent mRNA dose (Fig.6d). Example 7. Assessment of potential off-target edits To nominate potential genome-wide off-target sites, we performed the CHANGE-seq assay that captures gRNA / Cas-dependent DNA DSBs from a library of fragmented genomic DNA molecules (44). We used genomic DNA from patient- derived CYBB c.676C>T cells to ensure the detection of the on-target site as a positive control. (Fig.7a). Across replicate experiments using genomic DNA from two separate donors, we identified 32 potential off-target candidates with consistent off- target nomination between both genomic DNA samples (specificity ratio=0.75; Fig. 7a). CHANGE-seq experiments using the same patient genomic DNA but with the high fidelity variant SpRY-HF1 nuclease identified generally fewer off-targets and also substantially reduced unwanted editing of a prevalent off-target site with ABE8e SpRY on chromosome 4 (Fig.7a). The majority of the CHANGE-seq nominated sites are located in intergenic regions or introns and therefore less likely affect protein expression or function. To determine if the nominated sites exhibited bona fide off- target editing in HSPCs, we performed targeted multiplex amplicon sequencing of the on-target site and the top 50 CHANGE-seq nominated off-target sites using genomic DNA from CYBB c.676C>T HSCs treated with ABE8e-SpRY or ABE8e-SpRY-HiFi mRNA and gRNA A5 (Fig.7b, Figs.13A-B). With ABE8e-SpRY, we observed detectable off-target editing at approximately 20 of the 50 off-target sites (Fig.7b). However, when utilizing the the higher fidelity enzyme variant ABE8e-SpRY-HiFi, off-target editing was substantially reduced with much lower level off-target editing detected at only 6 off-target sites (Fig.7b and Figs.13A-B). Notably, 4 of the off- target sites were located in intergenic regions, and the other 2 were located in introns Attorney Docket No.29539-0835WO1 / MGH 2024-374 of the LAMB1 gene and a putative lncRNA (~250 bp and ~2kb from exons, respectively). Example 8. Whole Exome Sequencing Analysis Next, we performed whole exome sequencing (WES) to survey all protein- coding sequences in the human genome (45). The ability to achieve very high sequencing coverage via WES (500-1,000x at~$1,000) is superior compared to whole genome sequencing (30x at ~$10,000) since the exome comprises only about 1% of the genome (46). WES was performed on genomic DNA from two CYBB c.676C>T patient HSPCs treated with ABE8e-SpRY and gRNA A5 and their respective naïve untreated controls (Fig.14a). Our WES analysis achieved ~500x coverage for two different library preparation protocols that together provided a total of ~1,000x combined coverage for each treated sample, enabling detection of consistent off-target editing observed at efficiencies below 1% (at least 3 supporting reads in 500x mean coverage). To analyze the WES data, we first applied a pipeline designed specifically for low frequency variant calling (Agilent SureCall). Using human genome reference (hg38), around ~60,000 variant calls were made in all samples, including treated and untreated samples (Fig.14b). The majority of the variant calls were germline variants since they were present in the patient untreated naïve samples. After filtering out the germline variants from the treated sample, tens of thousands of variants remained in the treated sample (Fig.14b), as previously reported (47). However, over 80% of these private variants are known single nucleotide polymorphisms (SNPs). These are likely germline variants that failed to be filtered out and not true BE-derived OTs. While sensitive detection of potential off-target editing is critical for safety, false-positive variant calls can equally be detrimental. Single-sample variant calling fails to jointly account for variant quality and evidence in the control and edited samples, while germline style calling (e.g., GATK Best Practices) makes allele frequency assumptions that off-target variants will often violate. To more effectively balance sensitivity and precision, we used the genome-seek workflow (github.com / OpenOmics / genome-seek) to perform variant calling jointly in pairs of control and edited samples. This new pipeline removed the majority of the false positive mutation calls and confirmed only one mutation at chr12:57727421 in the AGAP2 gene that was absent in naïve but present in all 4 treated libraries (both CS Attorney Docket No.29539-0835WO1 / MGH 2024-374 and EF). The frequency of this variant, however, was low (3 / 751, 12 / 760, 3 / 722, and 2 / 755) in all treated samples. This may represent a bona fide BE-induced variant, or sequencing error / noise due to lower coverage (100-200x) in the naïve samples. Our WES analysis which detected almost no off-target edits clearly confirmed efficient on-target conversion of the mutant T allele to the WT C allele in treated samples (>80% VAF). The WES data also identified a bystander edit 2 bp upstream of target base as shown by NGS of the target site (Fig.7c). The bystander edit affected the intronic splice acceptor site in CYBB intron 6 (named i6-2) that was targeted in our HD model. Example 9. Analysis of potential transcriptome off-targets To evaluate potential off-target RNA edits that have been reported previously (48), a transcriptome analysis by RNA sequencing (RNA-seq) of CGD BE-HSPCs from 2 donors were compared with naïve unmanipulated HSPC samples. This transcriptome analysis did not identify any off-target edits at sites remote from the intended on-target site. At the target CYBB transcript, although most of the exon-exon junctions were normal, there was a low frequency of abnormal junctions of nearby exons (Ex5-Ex6, Ex7 to Ex8) to create Ex5 to Ex8 junctions. This exon-skipping resulted from a bystander edit on the splice site expected to abrogate gp91phox expression in those cells (~1%), which would effectively reverse the phenotypic impact of correction in ~1% of edited cells (Fig.7d). The RNA sequencing of the two CGD BE-HSPCs and naïve CGD HSPCs found no evidence of off-target RNA A-to- G edits compared to the naïve sample. The A-to-G variants that were observed were increased in chromosome 6 that corresponds to polymorphisms in the HLA, detected in both the BE and naïve. Example 10. Analysis of potential genomic structural alterations & genotoxicities To detect if our BE treatment caused large structural alterations, we performed conventional karyotyping, which did not detect any large chromosomal rearrangements in our HSPCs treated with ABE8e-SpRY and the CYBB c.676C>T A5 gRNA (Fig.8a). We then performed optical genome mapping (OGM) to detect genomic alterations using labeled long DNA molecules (>300 kb) to provide a high- resolution genome-wide analysis of large structural variants (SV). We established the Attorney Docket No.29539-0835WO1 / MGH 2024-374 sensitivity of OGM assay at below 5% VAF at 1.5 Tb coverage and 1% at 5 Tb by spiking in well characterized translocation Kasumi-1 leukemia cell line. Most of the SVs observed in the treated sample (Fig.8b) were filtered out after comparison with the naïve sample, leaving one SV. Upon manual examination, this single SV was also found in the naïve sample, indicating that no structural variants unique to the base- edited sample were identified within the detection limit of OGM (~1%) (Figs.8c). We also performed a single cell electrophoresis comet assay which detects DNA damage such as a loss of DNA supercoiling, which causes formation of a ‘comet tail’ during electrophoresis. Other factors such as cell membrane status may also affect comet formation. At 2 days after EP, CD34+HSPCs treated with positive control CRISPR-Cas9 nuclease and gRNA(11), or puromycin, a known cytotoxic agent, resulted in abundant comets (Fig.8d). Comets were also observed in CYBB c.676C>T HSPCs treated with ABE8e-SpRY and the A5 gRNA, although to a lesser extent than nuclease-treated cells and the abundance of comets decreased rapidly by day 3 and resolving by day 5 (Fig.8e). In contrast, Cas9 nuclease treated cells continued to display damage, except when treated with hu-genetic suppressor element (GSE)-56 (42) (49), while nearly all puromycin-treated cells exhibited comets (Fig.8e). The transient detection of comets in ABE-treated HSPCs may be in response to the nickase activity of the base editors or endonucleolytic repair following deamination. REFERENCES 1. S. S. De Ravin, X. Wu, S. Moir, S. Anaya-O'Brien, N. Kwatemaa, P. Littel, N. Theobald, U. Choi, L. Su, M. Marquesen, D. Hilligoss, J. Lee, C. M. Buckner, K. A. Zarember, G. O'Connor, D. McVicar, D. Kuhns, R. E. Throm, S. Zhou, L. D. Notarangelo, I. C. Hanson, M. J. Cowan, E. Kang, C. Hadigan, M. Meagher, J. T. Gray, B. P. Sorrentino, H. L. Malech, L. Kardava, Lentiviral hematopoietic stem cell gene therapy for X-linked severe combined immunodeficiency. Sci Transl Med 8, 335ra357 (2016). 2. D. B. Kohn, C. Booth, E. M. Kang, S. Y. Pai, K. L. Shaw, G. Santilli, M. Armant, K. F. Buckland, U. Choi, S. S. De Ravin, M. J. Dorsey, C. Y. Kuo, D. Leon-Rico, C. Rivat, N. Izotova, K. Gilmour, K. Snell, J. X. Dip, J. Darwish, E. C. Morris, D. Terrazas, L. D. Wang, C. A. Bauser, T. Paprotka, D. B. Kuhns, J. Gregg, H. E. Raymond, J. K. Everett, G. Honnet, L. Biasco, P. E. Newburger, F. D. Bushman, M. Grez, H. B. Gaspar, D. A. Williams, H. L. Malech, A. Galy, A. J. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Thrasher, C. G. D. c. Net, Lentiviral gene therapy for X-linked chronic granulomatous disease. Nat Med 26, 200-206 (2020). 3. L. Sereni, M. C. Castiello, D. Di Silvestre, P. Della Valle, C. Brombin, F. Ferrua, M. P. Cicalese, L. Pozzi, M. Migliavacca, M. E. Bernardo, C. Pignata, R. Farah, L. D. Notarangelo, N. Marcus, L. Cattaneo, M. Spinelli, S. Giannelli, M. Bosticardo, K. van Rossem, A. D'Angelo, A. Aiuti, P. Mauri, A. Villa, Lentiviral gene therapy corrects platelet phenotype and function in patients with Wiskott-Aldrich syndrome. J Allergy Clin Immunol 144, 825-838 (2019). 4. F. Ferrua, M. P. Cicalese, S. Galimberti, S. Giannelli, F. Dionisio, F. Barzaghi, M. Migliavacca, M. E. Bernardo, V. Calbi, A. A. Assanelli, M. Facchini, C. Fossati, E. Albertazzi, S. Scaramuzza, I. Brigida, S. Scala, L. Basso-Ricci, R. Pajno, M. Casiraghi, D. Canarutto, F. A. Salerio, M. H. Albert, A. Bartoli, H. M. Wolf, R. Fiori, P. Silvani, S. Gattillo, A. Villa, L. Biasco, C. Dott, E. J. Culme-Seymour, K. van Rossem, G. Atkinson, M. G. Valsecchi, M. G. Roncarolo, F. Ciceri, L. Naldini, A. Aiuti, Lentiviral haemopoietic stem / progenitor cell gene therapy for treatment of Wiskott-Aldrich syndrome: interim results of a non-randomised, open-label, phase 1 / 2 clinical study. Lancet Haematol 6, e239-e253 (2019). 5. A. O. Gupta, G. Raymond, E. I. Pierpont, S. Kemp, R. S. McIvor, A. Rayannavar, B. Miller, T. C. Lund, P. J. Orchard, Treatment of cerebral adrenoleukodystrophy: allogeneic transplantation and lentiviral gene therapy. Expert Opin Biol Ther 22, 1151-1162 (2022). 6. S. S. De Ravin, S. Liu, C. L. Sweeney, J. Brault, N. Whiting-Theobald, M. Ma, T. Liu, U. Choi, J. Lee, S. A. O'Brien, P. Quackenbush, T. Estwick, A. Karra, E. Docking, N. Kwatemaa, S. Guo, L. Su, Z. Sun, S. Zhou, J. Puck, M. J. Cowan, L. D. Notarangelo, E. Kang, H. L. Malech, X. Wu, Lentivector cryptic splicing mediates increase in CD34+ clones expressing truncated HMGA2 in human X-linked severe combined immunodeficiency. Nat Commun 13, 3710 (2022). 7. M. Cavazzana-Calvo, E. Payen, O. Negre, G. Wang, K. Hehir, F. Fusil, J. Down, M. Denaro, T. Brady, K. Westerman, R. Cavallesco, B. Gillet-Legrand, L. Caccavelli, R. Sgarra, L. Maouche-Chretien, F. Bernaudin, R. Girot, R. Dorazio, G. J. Mulder, A. Polack, A. Bank, J. Soulier, J. Larghero, N. Kabbara, B. Dalle, B. Gourmel, G. Socie, S. Chretien, N. Cartier, P. Aubourg, A. Fischer, K. Cornetta, F. Galacteros, Y. Beuzard, E. Gluckman, F. Bushman, S. Hacein-Bey-Abina, P. Attorney Docket No.29539-0835WO1 / MGH 2024-374 Leboulch, Transfusion independence and HMGA2 activation after gene therapy of human beta-thalassaemia. Nature 467, 318-322 (2010). 8. J. A. Winkelstein, M. C. Marino, R. B. Johnston, Jr., J. Boyle, J. Curnutte, J. I. Gallin, H. L. Malech, S. M. Holland, H. Ochs, P. Quie, R. H. Buckley, C. B. Foster, S. J. Chanock, H. Dickler, Chronic granulomatous disease. Report on a national registry of 368 patients. Medicine (Baltimore) 79, 155-169 (2000). 9. B. E. Marciano, C. Spalding, A. Fitzgerald, D. Mann, T. Brown, S. Osgood, L. Yockey, D. N. Darnell, L. Barnhart, J. Daub, L. Boris, A. P. Rump, V. L. Anderson, C. Haney, D. B. Kuhns, S. D. Rosenzweig, C. Kelly, A. Zelazny, T. Mason, S. S. DeRavin, E. Kang, J. I. Gallin, H. L. Malech, K. N. Olivier, G. Uzel, A. F. Freeman, T. Heller, C. S. Zerbe, S. M. Holland, Common severe infections in chronic granulomatous disease. Clin Infect Dis 60, 1176-1183 (2015). 10. B. E. Marciano, S. D. Rosenzweig, D. E. Kleiner, V. L. Anderson, D. N. Darnell, S. Anaya-O'Brien, D. M. Hilligoss, H. L. Malech, J. I. Gallin, S. M. Holland, Gastrointestinal involvement in chronic granulomatous disease. Pediatrics 114, 462-468 (2004). 11. S. S. De Ravin, L. Li, X. Wu, U. Choi, C. Allen, S. Koontz, J. Lee, N. Theobald-Whiting, J. Chu, M. Garofalo, C. Sweeney, L. Kardava, S. Moir, A. Viley, P. Natarajan, L. Su, D. Kuhns, K. A. Zarember, M. V. Peshwa, H. L. Malech, CRISPR-Cas9 gene repair of hematopoietic stem cells from patients with X-linked chronic granulomatous disease. Sci Transl Med 9, (2017). 12. M. L. Leibowitz, S. Papathanasiou, P. A. Doerfler, L. J. Blaine, L. Sun, Y. Yao, C. Z. Zhang, M. J. Weiss, D. Pellman, Chromothripsis as an on-target consequence of CRISPR-Cas9 genome editing. Nat Genet 53, 895-905 (2021). 13. K. Nishida, T. Arazoe, N. Yachie, S. Banno, M. Kakimoto, M. Tabata, M. Mochizuki, A. Miyabe, M. Araki, K. Y. Hara, Z. Shimatani, A. Kondo, Targeted nucleotide editing using hybrid prokaryotic and vertebrate adaptive immune systems. Science 353, (2016). 14. N. M. Gaudelli, A. C. Komor, H. A. Rees, M. S. Packer, A. H. Badran, D. I. Bryson, D. R. Liu, Programmable base editing of A*T to G*C in genomic DNA without DNA cleavage. Nature 551, 464-471 (2017). Attorney Docket No.29539-0835WO1 / MGH 2024-374 15. Y. B. Kim, A. C. Komor, J. M. Levy, M. S. Packer, K. T. Zhao, D. R. Liu, Increasing the genome-targeting scope and precision of base editing with engineered Cas9-cytidine deaminase fusions. Nat Biotechnol 35, 371-376 (2017). 16. A. C. Komor, Y. B. Kim, M. S. Packer, J. A. Zuris, D. R. Liu, Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature 533, 420-424 (2016). 17. T. P. Huang, G. A. Newby, D. R. Liu, Publisher Correction: Precision genome editing using cytosine and adenine base editors in mammalian cells. Nat Protoc 16, 5740 (2021). 18. E. M. Porto, A. C. Komor, I. M. Slaymaker, G. W. Yeo, Base editing: advances and therapeutic opportunities. Nat Rev Drug Discov 19, 839-859 (2020). 19. M. Jinek, K. Chylinski, I. Fonfara, M. Hauer, J. A. Doudna, E. Charpentier, A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337, 816-821 (2012). 20. B. P. Kleinstiver, M. S. Prew, S. Q. Tsai, V. V. Topkar, N. T. Nguyen, Z. Zheng, A. P. Gonzales, Z. Li, R. T. Peterson, J. R. Yeh, M. J. Aryee, J. K. Joung, Engineered CRISPR-Cas9 nucleases with altered PAM specificities. Nature 523, 481- 485 (2015). 21. W. Jiang, D. Bikard, D. Cox, F. Zhang, L. A. Marraffini, RNA-guided editing of bacterial genomes using CRISPR-Cas systems. Nat Biotechnol 31, 233-239 (2013). 22. R. T. Walton, K. A. Christie, M. N. Whittaker, B. P. Kleinstiver, Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science 368, 290-296 (2020). 23. Q. Ren, S. Sretenovic, S. Liu, X. Tang, L. Huang, Y. He, L. Liu, Y. Guo, Z. Zhong, G. Liu, Y. Cheng, X. Zheng, C. Pan, D. Yin, Y. Zhang, W. Li, L. Qi, C. Li, Y. Qi, Y. Zhang, PAM-less plant genome editing using a CRISPR-SpRY toolbox. Nat Plants 7, 25-33 (2021). 24. Y. Wu, Q. Ren, Z. Zhong, G. Liu, Y. Han, Y. Bao, L. Liu, S. Xiang, S. Liu, X. Tang, J. Zhou, X. Zheng, S. Sretenovic, T. Zhang, Y. Qi, Y. Zhang, Genome- wide analyses of PAM-relaxed Cas9 genome editors reveal substantial off-target effects by ABE8e in rice. Plant Biotechnol J 20, 1670-1682 (2022). Attorney Docket No.29539-0835WO1 / MGH 2024-374 25. J. Vicencio, C. Sanchez-Bolanos, I. Moreno-Sanchez, D. Brena, C. E. Vejnar, D. Kukhtar, M. Ruiz-Lopez, M. Cots-Ponjoan, A. Rubio, N. R. Melero, J. Crespo-Cuadrado, C. Carolis, A. J. Perez-Pulido, A. J. Giraldez, B. P. Kleinstiver, J. Ceron, M. A. Moreno-Mateos, Genome editing in animals with minimal PAM CRISPR-Cas9 enzymes. Nat Commun 13, 2601 (2022). 26. F. Liang, Y. Zhang, L. Li, Y. Yang, J. F. Fei, Y. Liu, W. Qin, SpG and SpRY variants expand the CRISPR toolbox for genome editing in zebrafish. Nat Commun 13, 3421 (2022). 27. M. F. Richter, K. T. Zhao, E. Eton, A. Lapinaite, G. A. Newby, B. W. Thuronyi, C. Wilson, L. W. Koblan, J. Zeng, D. E. Bauer, J. A. Doudna, D. R. Liu, Phage-assisted evolution of an adenine base editor with improved Cas domain compatibility and activity. Nat Biotechnol 38, 883-891 (2020). 28. B. P. Kleinstiver, V. Pattanayak, M. S. Prew, S. Q. Tsai, N. T. Nguyen, Z. Zheng, J. K. Joung, High-fidelity CRISPR-Cas9 nucleases with no detectable genome-wide off-target effects. Nature 529, 490-495 (2016). 29. C. A. Vakulskas, D. P. Dever, G. R. Rettig, R. Turk, A. M. Jacobi, M. A. Collingwood, N. M. Bode, M. S. McNeill, S. Yan, J. Camarena, C. M. Lee, S. H. Park, V. Wiebking, R. O. Bak, N. Gomez-Ospina, M. Pavel-Dinu, W. Sun, G. Bao, M. H. Porteus, M. A. Behlke, A high-fidelity Cas9 mutant delivered as a ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells. Nat Med 24, 1216-1224 (2018). 30. H. S. Kim, Y. K. Jeong, J. K. Hur, J. S. Kim, S. Bae, Adenine base editors catalyze cytosine conversions in human cells. Nat Biotechnol 37, 1145-1148 (2019). 31. K. A. Christie, B. P. Kleinstiver, Making the cut with PAMless CRISPR-Cas enzymes. Trends Genet 37, 1053-1055 (2021). 32. S. M. Miller, T. Wang, P. B. Randolph, M. Arbab, M. W. Shen, T. P. Huang, Z. Matuszek, G. A. Newby, H. A. Rees, D. R. Liu, Continuous evolution of SpCas9 variants compatible with non-G PAMs. Nat Biotechnol 38, 471-481 (2020). 33. C. A. Burnett, A. T. Wong, C. A. Vasquez, C. A. McHugh, G. W. Yeo, A. C. Komor, Examination of the Cell Cycle Dependence of Cytosine and Adenine Base Editors. Front Genome Ed 4, 923718 (2022). Attorney Docket No.29539-0835WO1 / MGH 2024-374 34. M. Dumont-Lagace, Q. Li, M. Tanguay, J. Chagraoui, T. Kientega, G. B. Cardin, A. Brasey, A. Trofimov, C. Carli, I. Ahmad, N. M. Bambace, L. Bernard, T. L. Kiss, J. Roy, D. C. Roy, S. Lemieux, C. Perreault, F. Rodier, S. F. Dufresne, L. Busque, S. Lachance, G. Sauvageau, S. Cohen, J. S. Delisle, UM171-Expanded Cord Blood Transplants Support Robust T Cell Reconstitution with Low Rates of Severe Infections. Transplant Cell Ther 27, 76 e71-76 e79 (2021). 35. J. Zeng, Y. Wu, C. Ren, J. Bonanno, A. H. Shen, D. Shea, J. M. Gehrke, K. Clement, K. Luk, Q. Yao, R. Kim, S. A. Wolfe, J. P. Manis, L. Pinello, J. K. Joung, D. E. Bauer, Therapeutic base editing of human hematopoietic stem cells. Nat Med 26, 535-541 (2020). 36. A. Hendel, R. O. Bak, J. T. Clark, A. B. Kennedy, D. E. Ryan, S. Roy, I. Steinfeld, B. D. Lunstad, R. J. Kaiser, A. B. Wilkens, R. Bacchetta, A. Tsalenko, D. Dellinger, L. Bruhn, M. H. Porteus, Chemically modified guide RNAs enhance CRISPR-Cas genome editing in human primary cells. Nat Biotechnol 33, 985-989 (2015). 37. P. Genovese, G. Schiroli, G. Escobar, T. D. Tomaso, C. Firrito, A. Calabria, D. Moi, R. Mazzieri, C. Bonini, M. C. Holmes, P. D. Gregory, M. van der Burg, B. Gentner, E. Montini, A. Lombardo, L. Naldini, Targeted genome editing in human repopulating haematopoietic stem cells. Nature 510, 235-240 (2014). 38. M. Mohrin, E. Bourke, D. Alexander, M. R. Warr, K. Barry-Holson, M. M. Le Beau, C. G. Morrison, E. Passegue, Hematopoietic stem cell quiescence promotes error-prone DNA repair and mutagenesis. Cell Stem Cell 7, 174-185 (2010). 39. L. W. Terstappen, S. Huang, M. Safford, P. M. Lansdorp, M. R. Loken, Sequential generations of hematopoietic colonies derived from single nonlineage-committed CD34+CD38- progenitor cells. Blood 77, 1218-1227 (1991). 40. P. M. Lansdorp, H. J. Sutherland, C. J. Eaves, Selective expression of CD45 isoforms on functional subpopulations of CD34+ hemopoietic cells from human bone marrow. J Exp Med 172, 363-366 (1990). 41. D. Wisniewski, M. Affer, J. Willshire, B. Clarkson, Further phenotypic characterization of the primitive lineage- CD34+CD38-CD90+CD45RA- hematopoietic stem cell / progenitor cell sub-population isolated from cord blood, mobilized peripheral blood and patients with chronic myelogenous leukemia. Blood Cancer J 1, e36 (2011). Attorney Docket No.29539-0835WO1 / MGH 2024-374 42. J. Brault, T. Liu, E. Bello, S. Liu, C. L. Sweeney, R. J. Meis, S. Koontz, C. Corsino, U. Choi, G. Vayssiere, M. Bosticardo, K. Dowdell, C. R. Lazzarotto, A. B. Clark, L. D. Notarangelo, J. C. Ravell, M. J. Lenardo, B. P. Kleinstiver, S. Q. Tsai, X. Wu, G. A. Dahl, H. L. Malech, S. S. De Ravin, CRISPR- targeted MAGT1 insertion restores XMEN patient hematopoietic stem cells and lymphocytes. Blood 138, 2768-2780 (2021). 43. S. Cohen, N. Bambace, I. Ahmad, J. Roy, X. Tang, M. J. Zhang, L. Burns, F. Barabe, L. Bernard, J. S. Delisle, T. Kiss, S. Lachance, D. C. Roy, O. Veilleux, G. Sauvageau, Improved outcomes of UM171-expanded cord blood transplantation compared with other graft sources: real-world evidence. Blood Adv 7, 5717-5726 (2023). 44. C. R. Lazzarotto, N. L. Malinin, Y. Li, R. Zhang, Y. Yang, G. Lee, E. Cowley, Y. He, X. Lan, K. Jividen, V. Katta, N. G. Kolmakova, C. T. Petersen, Q. Qi, E. Strelcov, S. Maragh, G. Krenciute, J. Ma, Y. Cheng, S. Q. Tsai, CHANGE-seq reveals genetic and epigenetic effects on CRISPR-Cas9 genome-wide activity. Nat Biotechnol 38, 1317-1327 (2020). 45. S. B. Ng, E. H. Turner, P. D. Robertson, S. D. Flygare, A. W. Bigham, C. Lee, T. Shaffer, M. Wong, A. Bhattacharjee, E. E. Eichler, M. Bamshad, D. A. Nickerson, J. Shendure, Targeted capture and massively parallel sequencing of 12 human exomes. Nature 461, 272-276 (2009). 46. A. J. Marian, Sequencing your genome: what does it mean? Methodist Debakey Cardiovasc J 10, 3-6 (2014). 47. M. Fiumara, S. Ferrari, A. Omer-Javed, S. Beretta, L. Albano, D. Canarutto, A. Varesi, C. Gaddoni, C. Brombin, F. Cugnata, E. Zonari, M. M. Naldini, M. Barcella, B. Gentner, I. Merelli, L. Naldini, Author Correction: Genotoxic effects of base and prime editing in human hematopoietic stem cells. Nat Biotechnol, (2024). 48. J. Grunewald, R. Zhou, S. P. Garcia, S. Iyer, C. A. Lareau, M. J. Aryee, J. K. Joung, Transcriptome-wide off-target RNA editing induced by CRISPR- guided DNA base editors. Nature 569, 433-437 (2019). 49. G. Schiroli, S. Ferrari, A. Conway, A. Jacob, V. Capo, L. Albano, T. Plati, M. C. Castiello, F. Sanvito, A. R. Gennery, C. Bovolenta, R. Palchaudhuri, D. T. Scadden, M. C. Holmes, A. Villa, G. Sitia, A. Lombardo, P. Genovese, L. Naldini, Attorney Docket No.29539-0835WO1 / MGH 2024-374 Preclinical modeling highlights the therapeutic potential of hematopoietic stem cell gene editing for correction of SCID-X1. Sci Transl Med 9, (2017). 50. L. W. Koblan, J. L. Doman, C. Wilson, J. M. Levy, T. Tay, G. A. Newby, J. P. Maianti, A. Raguram, D. R. Liu, Improving cytidine and adenine base editors by expression optimization and ancestral reconstruction. Nat Biotechnol 36, 843-846 (2018). 51. E. Ambrosino, J. A. Berzofsky, M. Terabe, Regulation of tumor immunity: the role of NKT cells. Expert Opin Biol Ther 8, 725-734 (2008). 52. Y. S. Slesarenko, A. V. Lavrov, S. A. Smirnikhina, Off-target effects of base editors: what we know and how we can reduce it. Curr Genet 68, 39-48 (2022). 53. H. A. Rees, C. Wilson, J. L. Doman, D. R. Liu, Analysis and minimization of cellular RNA editing by DNA adenine base editors. Sci Adv 5, eaax5717 (2019). 54. N. M. Gaudelli, D. K. Lam, H. A. Rees, N. M. Sola-Esteves, L. A. Barrera, D. A. Born, A. Edwards, J. M. Gehrke, S. J. Lee, A. J. Liquori, R. Murray, M. S. Packer, C. Rinaldi, I. M. Slaymaker, J. Yen, L. E. Young, G. Ciaramella, Directed evolution of adenine base editors with increased activity and therapeutic application. Nat Biotechnol 38, 892-900 (2020). 55. M. Fiumara, S. Ferrari, A. Omer-Javed, S. Beretta, L. Albano, D. Canarutto, A. Varesi, C. Gaddoni, C. Brombin, F. Cugnata, E. Zonari, M. M. Naldini, M. Barcella, B. Gentner, I. Merelli, L. Naldini, Genotoxic effects of base and prime editing in human hematopoietic stem cells. Nat Biotechnol, (2023). 56. K. Clement, H. Rees, M. C. Canver, J. M. Gehrke, R. Farouni, J. Y. Hsu, M. A. Cole, D. R. Liu, J. K. Joung, D. E. Bauer, L. Pinello, CRISPResso2 provides accurate and rapid genome editing sequence analysis. Nat Biotechnol 37, 224-226 (2019). 57. C. R. Lazzarotto, N. L. Malinin, Y. Li, R. Zhang, Y. Yang, G. Lee, E. Cowley, Y. He, X. Lan, K. Jividen, V. Katta, N. G. Kolmakova, C. T. Petersen, Q. Qi, E. Strelcov, S. Maragh, G. Krenciute, J. Ma, Y. Cheng, S. Q. Tsai, CHANGE-seq reveals genetic and epigenetic effects on CRISPR-Cas9 genome-wide activity. Nat Biotechnol, (2020). Attorney Docket No.29539-0835WO1 / MGH 2024-374 58. S. S. De Ravin, L. Li, X. Wu, U. Choi, C. Allen, S. Koontz, J. Lee, N. Theobald-Whiting, J. Chu, M. Garofalo, C. Sweeney, L. Kardava, S. Moir, A. Viley, P. Natarajan, L. Su, D. Kuhns, K. A. Zarember, M. V. Peshwa, H. L. Malech, CRISPR-Cas9 gene repair of hematopoietic stem cells from patients with X-linked chronic granulomatous disease. Sci Transl Med 9, eaah3480 (2017). 59. D. B. Kuhns, W. G. Alvord, T. Heller, J. J. Feld, K. M. Pike, B. E. Marciano, G. Uzel, S. S. DeRavin, D. A. Priel, B. P. Soule, K. A. Zarember, H. L. Malech, S. M. Holland, J. I. Gallin, Residual NADPH oxidase and survival in chronic granulomatous disease. N Engl J Med 363, 2600-2610 (2010). 60. H. Telenius, N. P. Carter, C. E. Bebb, M. Nordenskjold, B. A. Ponder, A. Tunnacliffe, Degenerate oligonucleotide-primed PCR: general amplification of target DNA by a single degenerate primer. Genomics 13, 718-725 (1992). 61. H. Telenius, A. H. Pelmear, A. Tunnacliffe, N. P. Carter, A. Behmel, M. A. Ferguson-Smith, M. Nordenskjold, R. Pfragner, B. A. Ponder, Cytogenetic analysis by chromosome painting using DOP-PCR amplified flow-sorted chromosomes. Genes Chromosomes Cancer 4, 257-263 (1992). 62. K. A. Christie, J. A. Guo, R. A. Silverstein, R. M. Doll, M. Mabuchi, H. E. Stutzman, J. Lin, L. Ma, R. T. Walton, L. Pinello, G. B. Robb, B. P. Kleinstiver, Precise DNA cleavage using CRISPR-SpRYgests. Nat Biotechnol 41, 409-416 (2023). 63. D. P. Cooke, D. C. Wedge, G. Lunter, Benchmarking small-variant genotyping in polyploids. Genome Res 32, 403-408 (2022). 64. S. Kim, K. Scheffler, A. L. Halpern, M. A. Bekritsky, E. Noh, M. Kallberg, X. Chen, Y. Kim, D. Beyter, P. Krusche, C. T. Saunders, Strelka2: fast and accurate calling of germline and somatic variants. Nat Methods 15, 591-594 (2018). 65. Y. Fan, L. Xi, D. S. Hughes, J. Zhang, J. Zhang, P. A. Futreal, D. A. Wheeler, W. Wang, MuSE: accounting for tumor heterogeneity using a sample- specific error model improves sensitivity and specificity in mutation calling from sequencing data. Genome Biol 17, 178 (2016). 66. B. S. Pedersen, J. M. Brown, H. Dashnow, A. D. Wallace, M. Velinder, M. Tristani-Firouzi, J. D. Schiffman, T. Tvrdik, R. Mao, D. H. Best, P. Bayrak-Toydemir, A. R. Quinlan, Effective variant filtering and expected candidate variant yield in studies of rare human disease. NPJ Genom Med 6, 60 (2021). Attorney Docket No.29539-0835WO1 / MGH 2024-374 67. M. F. Richter, K. T. Zhao, E. Eton, A. Lapinaite, G. A. Newby, B. W. Thuronyi, C. Wilson, L. W. Koblan, J. Zeng, D. E. Bauer, J. A. Doudna, D. R. Liu, Author Correction: Phage-assisted evolution of an adenine base editor with improved Cas domain compatibility and activity. Nat Biotechnol 38, 901 (2020). 68. Clinical Trial, Base Editing for Mutation Repair in Hematopoietic Stem & Progenitor Cells for X-Linked Chronic Granulomatous Disease; available at clinicaltrials.gov / study / NCT06325709 (March 21, 2024) OTHER EMBODIMENTS It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

Attorney Docket No.29539-0835WO1 / MGH 2024-374 WHAT IS CLAIMED IS:

1. A composition comprising a recombinant SpCas9 adenine base editor protein and a gRNA targeting CYBB, preferably as a ribonucleoprotein complex, wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

2. The composition of claim 1, wherein the recombinant SpCas9 adenine base editor protein and gRNA targeting CYBB are in a carrier, optionally wherein the carrier is a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle.

3. A composition comprising a nucleic acid sequence encoding an SpCas9 adenine base editor protein, and a gRNA targeting CYBB, wherein the gRNA comprises a sequence shown of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

4. A composition comprising a nucleic acid sequence encoding an SpCas9 adenine base editor protein and a nucleic acid sequence encoding a gRNA targeting CYBB,Attorney Docket No.29539-0835WO1 / MGH 2024-374 wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

5. The composition of claims 3 or 4, wherein the nucleic acid sequences encoding the base editor protein and / or gRNA are in a vector, e.g., a viral vector or a plasmid, and / or in a carrier, e.g., a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle.

6. A method of editing a cell comprising a mutation in an allele of CYBB, the method comprising contacting the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB, wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

7. An isolated cell from a subject who has a disease caused by a mutation in CYBB, wherein the mutation has been corrected by contacting the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB,Attorney Docket No.29539-0835WO1 / MGH 2024-374 wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

8. The isolated cell of claim 7, wherein the cell is a hematopoietic stem cell and progenitor cell (HSPC).

9. A method of treating a subject who has a disease caused by a mutation in CYBB, the method comprising administering to the subject a therapeutically effective amount of an isolated autologous cell, wherein the mutation in the cell has been corrected by contacting the cell with or expressing in the cell an SpCas9 adenine base editor protein and a gRNA targeting CYBB, wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

10. The method of claim 9, wherein the cell is a hematopoietic stem cell and progenitor cell (HSPC).Attorney Docket No.29539-0835WO1 / MGH 2024-374 11. A method of treating a subject who has a disease caused by a mutation in CYBB, the method comprising administering to the subject a therapeutically effective amount of a nucleic acid encoding an SpCas9 adenine base editor protein and a gRNA targeting CYBB, or a recombinant SpCas9 adenine base editor protein and a gRNA targeting CYBB, preferably as a ribonucleoprotein complex, wherein the gRNA comprises a sequence of one of SEQ ID NOs:1-49, optionally wherein: (a) the gRNA targets the BE to CYBB mutation c.676C>T and comprises A5 (SEQ ID NO:2), A4 (SEQ ID NO:1), or A3 (SEQ ID NO:3) gRNA, and the base editor comprises SpRY, SpRY-HF1, or SpRY-HiFi, preferably ABE8e- SpRY, ABE8e-SpRY-HF1, or ABE8e-SpRY-HiFi, or A7 (SEQ ID NO:4) with ABE8e-SpG or ABE8e-NRRH; or (b) the gRNA targets the BE to CYBB mutation c.1075G>A and comprises gRNA A6 (SEQ ID NO:5), A7 (SEQ ID NO:4), A8 (SEQ ID NO:6), and the base editor comprises WT SpCas9, preferably ABE8e-WT or ABE8e-WT- HF1, or A5 (SEQ ID NO:2) with ABE8e-SpG, or A9 (SEQ ID NO:7) with AB8e-SpRY.

12. The method of claim 11, wherein the nucleic acid sequences encoding the base editor protein and / or gRNA are in a vector, optionally a viral vector or a plasmid, and / or in a carrier, optionally a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle.

13. The method of claim 12, wherein the nucleic acid sequence encoding the base editor protein is an mRNA, and the method comprises administering the mRNA and gRNA in an LNP.

14. The method of claim 11, wherein the recombinant SpCas9 adenine base editor protein and gRNA targeting CYBB are administered in a carrier, optionally a lipid nanoparticle (LNP), liposome, exosome, or virus-like particle.