Targeted knock-in of NCF1 cDNA into the NCF2 locus leads to myeloid phenotypic correction of p47phox-deficient chronic granulomatous disease

p47phox-deficient chronic granulomatous disease (p47-CGD) is a primary immunodeficiency caused by mutations in the neutrophil cytosolic factor 1 (NCF1) gene, resulting in defective NADPH oxidase function in phagocytes. Due to its complex genomic context, the NCF1 locus is not suited for safe gene editing with current genome editing technologies. Therefore, we developed a targeted NCF1 coding sequence knock-in by CRISPR-Cas9 ribonucleoprotein and viral vector template delivery, to restore p47phox expression under the control of the endogenous NCF2 locus. NCF2 encodes for p67phox, an NADPH oxidase subunit that closely interacts with p47phox and is predominantly expressed in myeloid cells. This approach restored p47phox expression and NADPH oxidase function in p47-CGD patient hematopoietic stem and progenitor cells (HSPCs) and in p47phox-deficient mouse HSPCs, with the transgene expression following a myeloid differentiation pattern. Adeno-associated viral vectors performed favorably over integration-deficient lentiviral vectors for template delivery, with fewer off-target integrations and higher correction efficacy in HSPCs. Such myeloid-directed gene editing is promising for clinical CGD gene therapy, as it leads to the co-expression of p47phox and p67phox, ensuring spatiotemporal and near-physiological transgene expression in myeloid cells.


Supplemental Material
Supplemental Methods
In vitro cell-free cleavage assay PCR product (857 bp) containing the NCF2 on-target site was cloned into pCR Blunt II-TOPO vector, using Zero Blunt TOPO PCR Cloning Kit (Thermo Fisher Scientific, USA).The plasmid was linearized using SmaI (NEB, USA) as a target DNA for in vitro cleavage assay.Prior to RNP complexing, gRNA was denatured at 95 °C for 2 min and incubated at room temperature for 30 min.
RNP was complexed with 1.5 µM of Cas9-GFP protein and 1.5 µM of gRNA for 15 min at room temperature.55 ng of linearized target DNA was mixed with RNP complex and incubated for 1 hour at 37 °C.Reaction was stopped by addition of 0.02 M ethylenediamine tetraacetic acid (EDTA) and 1 µg/µL proteinase K (Roche, Switzerland), and incubated in a thermocycler (Biometra TOne, Analytik Jena, Germany) at 37 °C for 30 min.Digestion products were visualized on 1% agarose gel following electrophoresis.Band intensities were measured by densitometric analysis using ImageJ software (National Institutes of Health, USA). 1 Cleavage efficiency resulting from DSBs was calculated using the following formula: % Cleavage = 100 x (1 -(1-fraction cleaved) 1/2 ).

DNA isolation by Buffer K lysis
Genomic DNA from whole cells was lysed in Buffer K containing 0.1 mg/mL proteinase K (Roche, Switzerland), and 1X (v/v) GC buffer (Thermo Fisher Scientific, USA) in RNAse-/DNase-free water.
Samples were incubated at 57 °C for 45 min, followed by heat inactivation at 95 °C for 15 min.

T7 endonuclease I (T7EI) assay
Genomic DNA was isolated 4 days after treatment, PCR-purified using QIAquick Gel Extraction Kit or QIAquick PCR Purification Kit (QIAGEN, Germany) and subjected to T7EI assay.A total of 100 ng purified PCR products from untreated control and treated samples were hybridized in a thermocycler (Biometra TOne, Analytik Jena, Germany) by heating at 95 °C for 5 min, followed by slow cooling from 95 °C to 25 °C to form heteroduplexes.Heteroduplex DNA was then treated with 1U of T7EI (NEB, USA) and incubated in the thermocycler at 37 °C for 15 min.The reaction was blocked by addition of 0.02 M EDTA and analyzed by 1% agarose gel following electrophoresis.Band intensities were subjected to densitometric analysis using ImageJ software (National Institutes of Health, USA). 4 Non-homologous end joining (NHEJ) frequencies were calculated using the following formula: % Cleavage = 100 x (1 -(1-fraction cleaved) 1/2 ).
Copy number variation (CNV) analysis of NCF2 by ddPCR CNV quantification was determined by ddPCR using QX200 Droplet Reader (Bio-Rad Laboratories, USA).Genomic DNA was digested with DraI enzyme (20 U/µL) (NEB, USA).50-100 ng of digested gDNA was used per ddPCR reaction, primers, and probes (concentrations are stated in 'Materials and methods') targeting 3 locations upstream and 2 locations downstream of NCF2 target site using primer NCF2_T1 to _T5 with reference to FOXP2 housekeeping gene (Table S2).For droplet generation, 20 µL of sample reaction was transferred to a DG8 Cartridges for QX200/QX100 Droplet Generator (Bio-Rad Laboratories) and processed according to manufacturer's recommendation.Droplet generation and analysis were performed as described in the 'Materials and methods' section.

Characterization of identified integration sites by linear amplification mediated PCR (LAM-PCR)
The conditions for the first and second nested PCR were: initial denaturation for 10 min at 98 °C, then amplification with 35 cycles of denaturation (98 °C, 10 sec), annealing (62 °C, 10 sec) and extension (72 °C, 20 sec), with a final 5 min incubation at 72 °C.ISs were identified by Sanger sequencing of subcloned amplicons from the second nested PCR.The sequences were analysed for the presence of LTR III or ITR III primers, a restriction enzyme recognition site, and LC II primer.The unknown genome fragment was then subjected to standard nucleotide BLAST (blastn) using the National Center for Biotechnology Information (NCBI) webtool. 5To calculate the distance of IS from TSS, the closest TSS was determined by identification of the first nucleotide of the 5' untranslated region (5' UTR) of the gene with the IS.To verify if the IS contained potential gRNA binding sites, a list of CRISPR offtargets was generated in silico, using CRISPRoff (v1.2beta) and Cas-OFFinder, 6,7 with a query criteria of up to 6 mismatches, and 2 DNA or RNA bulges around the IS (200bp upstream and downstream).
The ISs were then manually analysed using the generated list of off targets, followed by a standard alignment using blastn.
Supplemental Tables Table S1.Guide RNA target sequences.Underlining denotes the PAM sequence of the sgRNA or the crRNA target.S1).(B) In vitro cell-free cleavage assay of Cas9 sgRNAs and Cas12a crRNAs.The % cleavage resulting from RNP activity is shown under the graph for each reaction.Different Cas-sgRNA targeting the NCF2 stop codon region were tested.Cas9 sgRNA #1 (subsequently referred to as sgRNA NCF2) was determined to have a high cleavage activity (45.8%) in an in vitro cell-free assay (Figure S1B) and in mammalian cells (39.0 ± 2.05%) by TIDE analysis (Figure S1E), with the Cas9 cleavage site located closest to the intended knock-in position (translational stop codon of NCF2).gated on both p67 phox positive and p67 phox negative populations that express p47 phox .All three replicates of cells treated with RNP + AAV NCF1 showed p47 phox expression that was mainly restricted to the p67 phox -positive population, with a minimal leakage into the p67 phox -negative population (up to 2.6%).
The gating of the p67        used during gene editing, the higher was the frequency of unintended integration events detected (unintended integrations mean integrations other than the on-target knock-in at NCF2).IDLVtransduced samples showed higher frequency of unintended integration events (e.g., at MOI 6x10 3 LP/cell, unintended integration events=14/24, knock-in efficiency of 35.6%) when compared with AAV-transduced samples (e.g., at MOI 1x10 7 vg/cell, unintended integration events=2/32, knock-in efficiency of 43.0%) even with higher knock-in efficiencies.
Figure S1.CRISPR-Cas gRNA design and screening in vitro and in cells.
(A) Position of Cas9 and Cas12a targeting the NCF2 locus at exon 15 and the 3' UTR.Red marks the stop codon of NCF2; grey line, target sequence; red line, PAM sequence; arrows, cleavage site (blunt/staggered cut) of the Cas9 or Cas12a target sequence (full list of gRNA sequences are shown in Table Figure 1D and 1E.(E) Left: indel frequency of respective gRNAs measured by TIDE analysis PLB-985 NCF1 ∆GT cells treated with Cas9 RNP (n=2).Right: results of the T7EI assay with the same samples (n=2 shown in 2 lanes).Asterisks mark bands of heteroduplex products cleaved by T7EI and cleavage efficiency in % is shown under the graph for each reaction.(F) Flow cytometry analysis of GFP expression in PLB-985 NCF1 ∆GT cells upon treatment with Cas9 sgRNA #1 RNP + IDLV GFP and Cas12a crRNA #7 RNP + IDLV GFP.Knock-in efficiency of SpCas9 RNP and AsCpf1/Cas12aRNP was compared.Although cleavage efficiency of Cas9 sgRNA #1 and Cas12a crRNA #7 was similar, the knock-in efficiency was low when the Cas12a system was used.NT, non-treated control.

Figure S2 .
Figure S2.Gating strategy of flow cytometry analyses in PLB-985 cells shown in Figure 1D.

Figure S3 .
Figure S3.Gating strategy of flow cytometry analyses in PLB-985 cells shown in Figure 1E.(A, C) Gating strategy for analysis of p47 phox expression in live, CD11b-positive cells.Gating of CD11b-positive cells is set on FMO control, with the addition of the isotype control PE-Cy7 antibody.Gating of the p47 phox -positive population within the CD11b-positive population is based on fully stained PLB-985 NCF1 ∆GT cells.Gating strategy for cells treated with RNP + IDLV NCF1 is presented in A, and for cells treated with RNP + AAV NCF1 in C. NT, non-treated control.(B) Representative flow cytometry showing cells treated with RNP + IDLV NCF1-GFP and RNP + IDLV NCF1 at the indicated MOIs.(D) Representative flow cytometry plots showing cells treated with RNP + AAV NCF1 at the indicated MOIs.(E) Relative p47 phox -positive MFI of corresponding samples in Figure 1E; n=2-3, data are shown as mean ± SD.MOI is denoted as LP/cell for IDLVs, or vg/cell for AAVs.

Figure S4 .
Figure S4.Gating strategy of flow cytometry analyses in PLB-985 cells shown in Figure 1F and 1G, and quantification of NBT test shown in Figure 1H (A) Gating strategy in Figure 1F to determine p47 phox positive cells in live, CD11b-positive population.FMO control plus isotype control PE-Cy7 antibody were used for gating of CD11b-positive cells, followed by gating of the p47 phox -positive population in fully stained PLB-985 NCF1 ∆GT cells.(B) Flow cytometry plots of p47 phox -positive populations in replicates 2 and 3 of corresponding samples in Figure 1F.Replicate 1 is shown as a representative plot in Figure 1F.(C) Gating strategy in Figure 1G to determine Rho-positive cells in live, CD11b-positive population.Gating of Rho-positive cells within the CD11b-positive population is set on fully stained, PMA-stimulated, non-treated PLB-985 NCF1 ∆GT cells as a negative control.(D) Flow cytometry plots of Rho-positive populations in replicates 2 and 3 of corresponding samples in Figure 1G.(E) Bar graphs show summary data of % p47 phox -positive cells and relative p47 phox -positive MFI gated on live, CD11b-positive cells of samples in Figure 1F.(F) Bar graphs show summary data of % Rho-positive cells and relative Rho positive MFI gated on live, CD11b-positive cells of samples in Figure 1G (In E and F, relative MFI is calculated as the ratio of MFI-positive population over the negative population of the same sample; n=3; ns, non-significant; data are shown as mean ± SD). (G) Quantification of the NBT test results in

Figure 1H .
Figure 1H.Statistical analysis performed with one-way ANOVA followed by Sidak's multiple comparisons test.

Figure S5 .
Figure S5.Gating strategy of flow cytometry analyses in PLB-985 cells shown in Figure 2A and 2B.(A) Flow cytometry analyses supplementary to Figure 2A to determine p67 phox and p47 phox expression in three independent replicates of PLB-985 WT cells, non-treated PLB-985 NCF1 ∆GT cells, and RNP + AAV NCF1 treated PLB-985 NCF1 ∆GT cells upon knock-in into NCF2.The plots show live cells, phox population was determined by FMO control, stained with isotype control AF488 antibody.(B) Gating strategy supplementary to Figure 2B to determine the p47 phox -positive population in undifferentiated cells and in myeloid-differentiated PLB-985 NCF1 ∆GT cells, gated on live and CD11b-positive cells.FMO controls, plus isotype control PE-Cy7 and APC antibodies were used to gate CD11b-and p47 phox -positive cells.NT, non-treated control.

Figure S6 .
Figure S6.IDLV knock-in in human CD34+ cells is limited by transduction efficiency and celltype tropism.

Figure S7 .
Figure S7.Gating strategy of flow cytometry analyses in lineage-negative mouse hematopoietic stem cells supplementary to Figure 3B, 3C, 3E and 3F.(A) Expression of p47 phox was gated on live, CD11b-positive cells.FMO control, stained with isotype control V450 antibody, was used for CD11b gating.Gating of p47 phox -positive cells was set on fully stained p47 phox -deficient mouse hematopoietic stem cells (mHSPCs) as negative control.The relative p47 phox -positive MFI of corresponding samples is shown in Figure 3B; n=2-3, data presented as mean

Figure S8 .
Figure S8.Editing of healthy human CD34+ hematopoietic stem cells shown in Figure 4A-D.

Figure S11 .
Figure S11.Flow cytometry analysis and DHR tests performed on 'Fused' and 'Cleaved' clones shown in Figure 5D.

Figure S12 .
Figure S12.Detailed characterization of individual clones knocked-in with 2A-NCF1 shown in Figure 6A.

Figure S13 .
Figure S13.Evaluation of copy number gains or losses surrounding the NCF2 on-target site.

Table S3 .
Oligonucleotide sequences for TIDE and Sanger sequencing.

Table S4 .
Antibodies and reagents for flow cytometry

Table S5 .
Information of integration sites identified by LAM-PCR in chosen knocked-in clones.Clone #, clone name described in FigureS12Cand S12D.VCN, vector copy number measured by ddPCR as described in 'Materials and methods'.IS chromosomal location, gene of IS,

Table S6 .
Primers used for LAM-PCR.Brackets indicate the specified modifications at the 5'.Underlined denotes TasI restriction enzyme recognition site.

Table S7 .
Sequence of integration sites identified by LAM-PCR in chosen knocked-in clones.Clone #, clone name described in FigureS11Cand S11D.(-) indicates genomic orientation of a reverse strand.Blue denotes the LTR III or ITR III primer sequence that binds to the LTR or ITR, respectively.Green denotes the linker LC II primer sequence.Underlined indicates the TasI restriction enzyme recognition site.Italicized indicates an unknown genomic sequence captured by LAM-PCR next to an LTR/ITR sequence.

Table S9 .
Primers used for NGS verification of CHANGE-seq off-target sites.

Table S10 .
Raw data of NGS deep sequencing for verification of CHANGE-seq off-target sites.

Table S11 .
Raw data of CHANGE-seq results.