Genome editing with the HDR-enhancing DNA-PKcs inhibitor AZD7648 causes large-scale genomic alterations

Abstract

The DNA-PKcs inhibitor AZD7648 enhances CRISPR-Cas9-directed homology-directed repair efficiencies, with potential for clinical utility, but its possible on-target consequences are unknown. We found that genome editing with AZD7648 causes frequent kilobase-scale and megabase-scale deletions, chromosome arm loss and translocations. These large-scale chromosomal alterations evade detection through typical genome editing assays, prompting caution in deploying AZD7648 and reinforcing the need to investigate multiple types of potential editing outcomes.

Fig. 1. AZD7648-induced HDR gain is skewed by an increased frequency of kilobase-scale deletions.

a, Schematic of editing workflow. K-562 cells were electroporated with a Cas9–sgRNA RNP with co-delivery of an ssODN HDR donor. After electroporation, edited cells were treated with 1 μM AZD7648 for 3 d and analyzed by short-read (Illumina) and long-read (ONT) sequencing. b, Frequency of HDR and indels detected by short-read sequencing in edited K-562 cells. c,d, Quantification by long-read sequencing of the fraction of total reads with a length greater than a defined size and quantified frequency of deletions larger than 1 kb in edited RPE-1 p53^−/− (c) and HSPCs (d). Data are from n = 1 biological replicate and n = 2 replicates with two distinct donors for the HSPC experiments.

Fig. 2. AZD7648 increases the frequency of megabase-scale deletions, chromosome arm loss and translocations.

a, Schematic of chromosome 7 around the eGFP insertion site in clonal K-562 cells, showing the eGFP location, −1.3-Mb editing site (located in SMURF1) and the +1-Mb editing site (located in IFT22). Cas9 RNP target sites are marked by red triangles. b, Quantification of the eGFP⁻ cell fraction detected by flow cytometry. c, K-562 eGFP cells edited at −1.3 Mb or with a non-targeting (NT) guide were analyzed by ddPCR to quantify copy numbers of the eGFP cassette in total edited cells. Copy number variation (CNV) is represented by the ratio of the eGFP copy number divided by the copy number detected at a control site on a different chromosome (chr3:46,270,956–46,271,086) and normalized to the unedited condition. In all conditions, gDNA was collected 6 d after electroporation. d, Schematic of the workflow for editing upper airway organoids. e, Heatmap of residual gene expression of GAPDH UTR-edited upper airway organoids. f, Frequency of upper airway organoid cells exhibiting gene expression loss in the 6.5-Mb telomeric or 7.0-Mb centromeric segment. Cells with average residual gene expression lower than 0.95 were considered as cells with gene expression loss. g, Circos plots of CAST-seq-detected chromosomal translocations in edited HSPCs. h, Schematic of the workflow for editing RPE-1^−/− cells with AZD7648 and PolQi2 in combination. i, Quantification by long-read sequencing of the fraction of total reads with a length greater than a defined size and quantified frequency of deletions larger than 1 kb in edited RPE-1 p53^−/−. Data are from n = 1 biological replicate (with three technical replicates for c), except for b with n = 4 biological replicates. b,c, Results are presented as mean ± s.d.

Extended Data Fig. 1. AZD7648-induced DNA-PKcs inhibition increases detected HDR in RPE-1 and primary cells.

(a-d) Frequency of HDR and indels detected by short-read sequencing in edited RPE-1 p53^−/− (a), RPE-1 p53^+/+ (b), HSPC (c), upper airway organoid cells (d). Genomic DNA was collected 3 d post-electroporation. Cells were electroporated with Cas9 in complex with a single guide RNA (sgRNA) as an RNP, with co-delivery of a single-stranded oligodeoxynucleotide (ssODN) HDR donors. Edited cells were treated with 1 μM AZD7648 for 3 d immediately following electroporation. Genomic DNA was collected 3 d post-electroporation. Data are from n = 1 biological replicate and one donor for HSPC experiments.

Extended Data Fig. 2. The FIRE reporter enables HDR and frame-shift indels quantification by flow cytometry.

(a) Schematic of the FIRE reporter construct. The FIRE reporter system consists of an in-frame disrupted mScarlet sequence (with a 228 nt deletion), followed by two out-of-frame eGFP cassettes (+1 bp and +2 bp). K-562 cells with the FIRE reporter were electroporated to induce a DSB in the disrupted mScarlet sequence. The HDR template is a long 528 nt ssODN containing the missing 228 nt of the mScarlet sequence flanked by two 150 nt-homology arms. HDR-mediated repair using the corrective ssODN is expected to lead to the mScarlet expression, while generation of indels causing frameshifts should trigger eGFP expression. Unmodified target sites, in-frame small indels, reporter silencing, and larger deletions that lead to loss of the reporter expression are expected to result in non-fluorescent cells. (b, c) Flow cytometry data of edited K-562 FIRE reporter cells (b) and quantification of HDR (mScarlet+) and indels (eGFP+) event frequencies (c) detected by flow cytometry (n = 4 biological replicates). (d) Quantification of HDR and indel reads detected by Sanger sequencing and Inference of CRISPR Edits (ICE) in edited K-562 cells at the FIRE reporter (n = 1 biological replicate). (e) Indel profile in Cas9-edited K-562 FIRE reporter cells assessed by ICE analysis. Analyses were performed 3 days post-electroporation for both flow cytometry and Sanger sequencing assays.

Extended Data Fig. 3. AZD7648 increases frequency of kilobase-scale deletions.

(a) Schematic of long-read deletion quantification. Reads obtained from ONT sequencing were initially quality filtered and aligned to the target site (see Methods). The fraction of total reads of a greater length than a defined size was calculated. A population of reads with deletions leads to a leftward shift of the curve. (b) AZD7648 induces kilobase-scale deletions in K-562 cells at the FIRE reporter. Raw data is shown in the left panel, and the quantified frequency of deletions larger than 1 kb is shown in the right panel. (c) Comparison of detected HDR events using three different readouts in K-562 cells with FIRE reporter edited with Cas9 + ssODN only (light blue) or treated with AZD7648 (dark blue). (d–f) Quantification of the fraction of total reads with a length greater than a defined size (top panel) and frequency of deletions larger than 1 kb, 2 kb or 3 kb (bottom panel) in edited RPE-1 p53^−/− (d) RPE-1 p53^+/+ (e) and K-562 (f). In all conditions, genomic DNA was collected 3 d post-electroporation. Data are from n = 1 biological replicate.

Extended Data Fig. 4. AZD7648 increases frequency of megabase-scale deletions and chromosome arm loss.

(a) Frequency of HDR and indels detected by short-read sequencing in K-562 eGFP cells edited at −1.3 Mb or +1 Mb relative to the reporter. (b, c) K-562 eGFP cells edited at −1.3 Mb were analyzed by droplet digital PCR to quantify copy numbers of the eGFP cassette in fractions sorted for eGFP+ and eGFP- cells with gating strategy shown in (c). (d) Schematic representation of chromosome 7 of −1.3 Mb-edited K-562 eGFP clonal cell line, illustrating potential outcomes editing at −1.3 Mb (e) K-562 eGFP cells edited at −1.3 Mb or with a non-targeting guide (NT) were analyzed by droplet digital PCR to quantify copy numbers of KMT2C. Edited cells were analyzed 6 d post-electroporation. Copy number variation (CNV) is represented by the ratio of eGFP (b) or KMT2C (e) copy numbers divided by the copy number detected at chr3:46,270,956-46,271,086, normalized to the unedited condition. Data are from n = 1 biological replicate with three technical replicates in (b) and (e). (b, e) Results are presented as mean ± standard deviation.

Extended Data Fig. 5. AZD7648 increases the frequency of gene expression loss on the targeted chromosome arm in primary cells and organoids.

(a, b) Flow cytometry data (a) and quantification (b) of cell viability of GAPDH UTR-edited (blue) or mock-electroporated (grey) HSPCs. The fraction of live cells is calculated as the ratio of SYTOX Red-negative cells to the total number of events analyzed. (c) Heatmap of residual gene expression of UTR-edited upper airway organoids across the entire genome. (d, e) Mean normalized gene expression of edited upper airway organoid cells (d) and frequency of upper airway organoid cells with gene expression gain (average residual gene expression > 1.05 – in red), unmodified gene expression (from 1.05 to 0.95), or gene expression loss (< 0.95 – in blue) (e) in the telomeric or centromeric segment of the GAPDH UTR target site. (f, g) Heatmap showing the residual gene expression of UTR-edited HSPCs across the entire genome (f) and specifically on chromosome 12 (g). (h, i) Mean normalized expression of transcripts (h) and frequency of HSPCs with gene expression gain (average residual gene expression > 1.05 – in red), unmodified gene expression (from 1.05 to 0.95), or gene expression loss (< 0.95 – in blue) (i) in the telomeric or centromeric segment of the GAPDH target site for edited HSPCs. The number of analyzed cells per condition in (c) and (f) ranged from 9,296 to 14,871. The red triangles indicate the Cas9 RNP target sites. Telomeric and centromeric segments were defined as the chromosome start to the GAPDH target site (chr12:6,535,922) and the 7.0 Mb region immediately downstream of the cut site. Data are from n = 1 biological replicate.

Extended Data Fig. 6. AZD7648 increases the frequency of translocations.

(a) Circos plots of detected chromosomal translocations in RPE-1 p53^−/− (left panel) and RPE-1 p53^+/+ (right panel) edited at HBB. On-target aberrations are shown in green, off-target mediated translocations (OMTs) in red, homology-mediated translocations (HMTs) in blue, and mixed types in yellow. (b) Quantification of translocation events. Shown is the number of reads (counts per million, CPM) for HBB OMT-1 (chr.9) in RPE-1 p53^−/− and HSPCs edited at HBB with or without AZD7648 and for CCR5 OMT-1 (chr. 11) in HSPCs edited at HBB with or without AZD7648 normalized by the total number of reads. (c) Circos plots of detected chromosomal translocations in RPE-1 p53^−/− (left panel) and RPE-1 p53^+/+ (right panel) edited at CCR5. (d) Quantification of translocation events. Indicated is the number of reads (counts per million, CPM) for OMT-1 (chr.14), OMT-2 (chr.11) and OMT-3 (chr.13) in CCR5-edited RPE-1 p53^−/− or RPE-1 p53^+/+ with AZD7648. Data are from n = 1 biological replicate with n = 2 technical replicates, except for RPE-1 p53^+/+ edited at CCR5 (n = 1).

Extended Data Fig. 7. Addition of PolQi2 preserves high HDR and limits occurrence of AZD7648-induced kilobase-scale deletions.

(a) Schematic of the editing workflow. RPE-1 p53^−/− and RPE-1 p53^+/+ cells were electroporated with a Cas9 RNP, with co-delivery of a ssODN HDR donor. Edited cells were treated with 1 μM AZD7648 and/or 3 μM PolQi2 for 3 d immediately following electroporation. (b, c) Frequency of HDR and indels detected by short-read sequencing in RPE-1 p53^−/− (b) and RPE-1 p53^+/+ cells (c) edited at HBB, CCR5 and GAPDH. Genomic DNA was collected 3 d post-electroporation. (d–f) Quantification by long-read sequencing of the fraction of total reads with a length greater than a defined size and quantified frequency of deletions larger than 1 kb in RPE-1 p53^−/− edited without the ssODN HDR donor (d) and RPE-1 p53^+/+ edited without (e) or with the ssODN HDR donor (f). Data are from n = 1 biological replicate.

Extended Data Fig. 8. AZD7648 and PolQi2 in combination did not reduce chromosome arm loss when co-delivered with an ssODN HDR template.

(a) Schematic of the region of chromosome 7 surrounding the eGFP insertion site in a clonal K-562 cell line. The position of the single-copy eGFP insertion site and of the SMURF1 (−1.3 Mb) locus are indicated. The red triangle represents the Cas9 RNP target site. (b) Frequency of HDR and indels detected by short-read sequencing in K-562 eGFP cells edited at SMURF1. Genomic DNA was collected 6 d post-electroporation. (c) Quantification by long-read sequencing of the fraction of total reads with a length greater than a defined size and quantified frequency of deletions larger than 1 kb in K-562 eGFP edited at SMURF1 in the absence of ssODN delivery (left panel) and in its presence (right panel). (d, e) Quantification (d) and illustrative flow cytometry data (e) of the eGFP-negative cell fraction. Edited cells were analyzed 6 d post-electroporation. Data are from n = 1 biological replicate, except for (d) with n = 2 biological replicates.