
==== Front
J Biol Chem
J Biol Chem
The Journal of Biological Chemistry
0021-9258
1083-351X
American Society for Biochemistry and Molecular Biology

S0021-9258(24)02196-3
10.1016/j.jbc.2024.107695
107695
Research Article
Cell cycle arrest combined with CDK1 inhibition suppresses genome-wide mutations by activating alternative DNA repair genes during genome editing
Fukuda Nozomi 1‡
Soga Keisuke 1‡
Taguchi Chie 1
Narushima Jumpei 1
Sakata Kozue 1
Kato Reiko 1
Yoshiba Satoko 1
Shibata Norihito 1
Kondo Kazunari k-kondo@swu.ac.jp
12∗
1 Division of Biochemistry, National Institute of Health Sciences, Kawasaki, Kanagawa, Japan
2 Faculty of Food and Health Sciences, Showa Women’s University, Tokyo, Japan
∗ For correspondence: Kazunari Kondo k-kondo@swu.ac.jp
‡ These authors contributed equally to this work.

17 8 2024
9 2024
17 8 2024
300 9 1076959 3 2024
8 8 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by/4.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Cells regularly repair numerous mutations. However, the effect of CRISPR/Cas9-induced dsDNA breaks on the repair processes of naturally occurring genome-wide mutations is unclear. In this study, we used TSCE5 cells with the heterozygous thymidine kinase genotype (TK+/−) to examine these effects. We strategically inserted the target sites for guide RNA (gRNA)/Cas9 and I-SceI into the functional allele and designed the experiment such that deletions of > 81 bp or base substitutions within exon five disrupted the TK gene, resulting in a TK−/− genotype. TSCE5 cells in the resting state exhibited 16 genome-wide mutations that affected cellular functions. After gRNA/Cas9 editing, these cells produced 859 mutations, including 67 high-impact variants that severely affected cellular functions under standard culture conditions. Mutation profile analysis indicated a significant accumulation of C to A substitutions, underscoring the widespread induction of characteristic mutations by gRNA/Cas9. In contrast, gRNA/Cas9-edited cells under conditions of S∼G2/M arrest and cyclin-dependent kinase 1 inhibition showed only five mutations. Transcriptomic analysis revealed the downregulation of DNA replication genes and upregulation of alternative DNA repair genes, such as zinc finger protein 384 (ZNF384) and dual specificity phosphatase, under S∼G2/M conditions. Additionally, activation of nucleotide and base excision repair gene, including O-6-methylguanine-DNA methyltransferase and xeroderma pigmentosum complementation group C, was observed. This study highlights the profound impact of CRISPR/Cas9 editing on genome-wide mutation processes and underscores the emergence of novel DNA repair pathways. Finally, our findings provide significant insights into the maintenance of genome integrity during genome editing.

Keywords

CRSIPR/Cas
genome-wide mutation
alternative DNA repair
Abbreviations

BAM binary alignment map

BER base excision repair

CDC cell division cycle

CDK1 cyclin-dependent kinase 1

CNV copy number variation

DONSON downstream neighbor of Son

d-thy double thymidine block

DSB double-strand break

DUSP dual specificity phosphatase

GO gene ontology

gRNA guide RNA

HR homologous recombination

MCM minichromosome maintenance

MF mutant frequency

MGMT O-6-methylguanine methyltransferase

MMR mismatch repair

NER nucleotide excision repair

NGS next-generation sequencing

NHEJ nonhomologous end joining

Non-syn nonsynchronous

RPE retinal pigment epithelial

SV structural variant

TFT trifluorothymidine

TK thymidine kinase

VEP variant effect predictor

XPC xeroderma pigmentosum complementation group C

ZNF zinc finger

Reviewed by members of the JBC Editorial Board. Edited by Patrick Sung
==== Body
pmcThe CRISPR/CRISPR-associated proteins (Cas) system has revolutionized genome engineering. This technology has been now extensively applied to the development of plant traits and gene therapies. The foundational CRISPR/Cas9 system from Streptococcus pyogenes comprises a guide RNA (gRNA) and the SpCas9 protein (1, 2). The gRNA identifies a 20 bp protospacer sequence that is complementary to the DNA target, whereas the Cas9 protein recognizes a three-base protospacer adjacent motif sequence on the opposite strand. This interaction forms an R-loop, allowing the gRNA–Cas9 complex to catalytically cleave the dsDNA at the specified site (3, 4, 5). Most double-strand breaks (DSBs) in cells are fixed by DNA repair mechanisms, such as nonhomologous end joining (NHEJ), microhomology-mediated end joining, and homology-directed repair. However, some DSBs can be restored by nucleotide deletions, insertions, or substitutions, which may result in gene knockouts.

Following the initial development of CRISPR/SpCas9, subsequent studies aimed to enhance SpCas9 fidelity, leading to the creation of high-specificity variants, such as SpCas9-HF (6) and HiFi-Cas9 (7). These variants have demonstrated precise editing capabilities in various cell lines (8). Another significant advancement was the discovery of the Cas12a (Prevotella and Francisella 1, Cpf1) ortholog, which exhibits a different protospacer adjacent motif recognition and cleavage pattern from those of SpCas9. Furthermore, Cpf1 generally produces precise edits with staggered ends (9, 10, 11, 12, 13).

Despite these efforts, SpCas9 remains the preferred tool for genome editing owing to its well-characterized effects on on-target and off-target sites. However, the off-target activity of CRISPR/Cas systems poses a significant concern, particularly for therapeutic applications, which have prompted extensive mitigating studies (14, 15, 16, 17, 18, 19). Although the high-fidelity Cas9 variants show improvements, they exhibit variable efficiencies and residual off-target activities, which could limit their clinical utility. Moreover, studies indicates that DNA repair outcomes are influenced by the protospacer sequence and are consistent across different cell lines, suggesting a nonrandom, context-dependent repair process (20). Deletion lengths range from a few to 20 bases (21, 22). The NHEJ pathway is the primary mechanism for DNA repair, which act immediately to inhibit large deletions and functions prominently even during the G2 phase when homologous recombination (HR) is active (23). A example of complex DNA breaks or chromatin rearrangement repair occurs if NHEJ cannot rapidly rejoin DNA ends, which initiates end resection and HR repair (24). Notably, larger deletions occur at low frequencies but are more pronounced in cells deficient in NHEJ pathway components, such as DNA ligase IV (LIG4), XRCC4, DNA-PKcs, and XLF. Conversely, deficiencies in Nijmegen breakage syndrome 1 (part of the MRE11–RAD50–Nijmegen breakage syndrome 1 complex) and DNA polymerase theta (POLQ) result in reduced deletion frequencies at the targeted sites (23, 25). These observations suggest that larger deletions tend to increase with the inhibition of NHEJ pathway components. DNA breaks induced in the G1 phase have been reported to be repaired in the S–G2/M phase, which leads to translocations (26).

To address the concerns surrounding the off-target effects of CRIPR/Cas9 editing, in this study, we leverages the unique characteristics of TK6 cells, which can be selected with trifluorothymidine (TFT) because of their ability to sustain deletions larger than 81 bp or substitutions within exon 5 (27). We hypothesized that frequent DNA cleavage by gRNA/Cas9 impairs the ability of cells to repair naturally occurring mutations, potentially leading to erroneous repairs or genomic instability. By synchronizing cell cycles and inhibiting cyclin-dependent kinase 1 (CDK1), we demonstrated that it is possible to mitigate the development of genome-wide mutations and rearrangements induced by genome editing. Furthermore, we explored the role of alternative DNA repair pathways such as those involving zinc finger (ZNF) proteins in suppressing genome instability during CRISPR-Cas editing.

Results

Lack of repair through NHEJ pathway increases genome editing efficiency at target site

Cells repair naturally occurring mutations every single day. Most mutations are correctly repaired in resting cells. However, we do not know how the induction of artificial DNA breaks affects the repair of mutations that occur naturally throughout the genome. When artificial DNA DSBs are induced by gRNA3/Cas9, cells begin to repair not only naturally occurring mutations but also genome-edited DNA cleavages (Fig. 1A). Do cells have the ability to correct all these mutations? If not, the cells will accumulate mutations and we need to find the best way to prevent them from the development of those mutations.Figure 1 The surviving cells, with disrupted exons in the allele A of TK1 gene, were analyzed to reveal genome-wide mutations.A, relationship between naturally occurring and genome-edited mutations. Cells repair naturally occurring mutations daily. Cells correct artificial DNA cleavage induced by genome editing, such as gRNA/Cas9. Some cells correct both naturally occurring and genome-edited mutations, whereas others repair one or neither. Consequently, genome-edited cells may contain both the desired mutation at the target site of the genome editing and naturally occurring mutations throughout the genome. We wondered whether genome editing increases naturally occurring mutations that need to be repaired. B, TSCE5 cells were constructed by inserting a 32-base sequence containing I-SceI recognition site (dotted square) at the 81 bp upstream of exon 5 (allele A) of TK6 cells, which have nonfunctional exon 4 in the other allele (allele B). A single guide RNA (gRNA3) was designed between the insertion sequence and the genome of TSCE5 cells. Large deletion involving exon 5 or other exons induced by gRNA3/Cas9 or I-SceI results in functional loss of TK1 and trifluorothymidine (TFT)-resistant phenotype. Substitutions within exon 5 also disrupts TK1. C, double-strand breaks (DSBs) was induced at each target site (I-SceI or gRNA3/Cas9). Cells were seeded into 96-well culture plates after a 3-day culture and then selected in the presence of TFT. Cells with TFT-resistant phenotypes can survive, whereas other cells are dead (empty wells). Genomic DNA (gDNA) was extracted from surviving cells (TFT-resistant) with disrupted exons in allele A of TK1 for whole-genome sequence analysis. Cell culture in 96-well culture plates generated large or small colonies. D, similar experiments were performed under synchronous (S–G2/M arrest followed by cyclin-dependent kinase 1 (CDK1) inhibition [S–G2/M] conditions, where double thymidine (d-thy) block and colcemid were used to synchronize cells before genome editing. CDK inhibitor Ro3306 (Ro) was also used to block cell division during genome editing by I-SceI or gRNA3/Cas9 addition (S–G2/M conditions), and cells that survived TFT selection formed one colony. TK, thymidine kinase.

To address this, we analyzed genome sequences of nonedited cells and edited cells by CRISPR/Cas9. In particular, we focused the genome-edited cells with larger deletions at the target site or with base substitutions in exon 5 because those mutations are derived from lack of the repair through the pathways, such as NHEJ and base/nucleotide excision repair (BER/NER). In this study, we used the cell system consisting of TSCE5 cells (28), with a heterozygous thymidine kinase (TK) genotype (TK+/−, Fig. 1B), which survived in the presence of TFT, to acquire cells harboring larger deletions (>81 bp) around the target site or base substitutions within exon 5. Target sites for gRNA3/Cas9 and I-SceI were inserted into the functional allele of TK1 gene (allele A), 81 bp before the exon 5, so that the formation of larger deletions or the occurrence of mutations within exon 5 disrupts the TK1 gene, which leads from TK+/-to TK−/− genotype. Methodologically, TFT is converted to TFT monophosphate by the TK1 gene in TK+/− cells, which inhibits thymidylate synthase for DNA synthesis. On the other hand, in TK−/− cells, TFT is not converted to TFT monophosphate. Therefore, TK−/− cells can survive in the presence of TFT. We then scrutinized the genomes of the surviving cells (Fig. 1, C and D). Whole genomes of the surviving cells were analyzed in detail using next-generation sequencing (NGS). We used three different cell lines. TSCE5 cells were used to analyze mutations derived from the NHEJ pathway. TSCE2 and 122 cells were used for measuring the ratios of DNA repair through the HR pathway. TSCE122 lacks whole exon 5 in TK1 gene, whereas TSCE2 has a point mutation in the exon 5. The repair of whole exon 5 is more difficult than correcting a single nucleotide substitution (Fig. S1). First, we measured the mutant frequency and the repair frequency of gRNA3/Cas9-edited and I-SceI–edited cells using the three different cells. In TSCE5 cells, two cell culture conditions were used: nonsynchronous (hereinafter referred to as non-syn) and synchronous conditions (S∼G2/M arrest followed by CDK1 inhibition, S∼G2/M). Mutant frequencies (MFs) are the occurrence ratios of larger deletions (>81 bp) around the target site or small mutations such as base substitutions within exon 5 of TK1. The MFs were 3.39 × 10−5 and 5.84 × 10−5 in gRNA3/Cas9-edited cells and I-SceI–edited cells, respectively, under non-syn conditions. In contrast, the MFs were 27.2 × 10-5 and 8.10 × 10−5 in gRNA3/Cas9-edited cells and I-SceI–edited cells, respectively, under S ∼ G2/M conditions (Fig. 2A). The frequency of gRNA3/Cas9-edited cells increased 8.02-fold under S∼G2/M conditions compared to non-syn conditions. This result is consistent with reports published so far (29, 30, 31, 32, 33). The results of the measurement of repair frequencies, which are the ratios of DNA repair through the HR pathway, in TSCE2 cells showed values of 0.47 × 10−5 and 0.80 × 10−5 in gRNA3/Cas9-and I-SceI–edited cells, respectively. In TSCE122 cells, the repair of whole exon 5 was detected only in I-SceI–edited cells with a repair frequency of 0.27 × 10−5. This is probably due to the difference in the DSB pattern of DNA, that is, blunt ends for gRNA3/Cas9 and staggered ends for I-SceI. Alternatively, the longer binding at the target site of gRNA3/Cas9 after DNA cleavage can be another cause.Figure 2 Thymidine kinase assay was used to selects cells with deletions > 81 bp or mutations in exon 5 using trifluorothymidine in TSCE5 cells, and to select cells repaired by homologous recombination using cytidine and HAT in TSCE2 and TSCE122 cells. TSCE2 carries 1 bp mutation in exon 5, while TSCE122 lacks 335 bp, including whole exon 5. A, mutations in TSCE5 cells detected in nonhomologous end-joining (NHEJ) pathway. Mutant frequencies (MFs) of large deletions (>81 bp) were almost the same in gRNA3/Cas9- and I-SceI–edited [MF = 3.39 versus 5.84 (×10−5)] under nonsynchronous (non-syn) conditions. In contrast, MF in gRNA3/Cas9-edited cells was much higher than that in I-SceI–edited [27.2 versus 8.10 (×10−5)] cells under S–G2/M conditions. gRNA/Cas9-edited cells increased the frequency 8.02-fold under S∼G2/M conditions compared to non-syn conditions. Data are from three independent experiments (mean ± SD). B, frequencies of HR against 1 bp point mutation (G- > A) at 23rd base from beginning of exon 5 and missing 335 bp were estimated in TSCE2 and TSCE122 cells under non-syn conditions. HR frequencies against 1 bp substitution were similar in I-SceI- and gRNA3/Cas9-edited cells. Repair of missing 335 bp was detected in I-SceI-, but not gRNA3/Cas9-edited cells. All 384 wells (four 96-well plates) were empty after drug selection in TK assay, indicating difficulty of gRNA/Cas9-initiated repair in HR pathway. Data are from three independent experiments (mean ± SD). gRNA, guide RNA.

Genome editing increased genome-wide mutations, which are suppressed under S–G2/M conditions

When DNA DSBs are induced by gRNA3/Cas9, cells need to correct the additional DNA lesions. To investigate the effect of DNA DSBs by gRNA3/Cas9 on the repair of naturally occurring mutations, we analyzed the genomes of nonedited TSCE5 cells, gRNA3/Cas9-edited TSCE5 cells and I-SceI–edited TSCE5 cells and compared the whole sequences of genome-edited cells with the sequence of nonedited cells to extract newly emerged mutations. Mutations such as base substitutions and indels were also investigated genome-wide and extracted using the GATK/Mutect2 software (Broad Institute, https://gatk.broadinstitute.org/hc/en-us). The mutations that occurred only in gRNA3/Cas9 or I-SceI–edited cells are summarized in Figure 3, Tables S1, and S2, showing cellular functional impacts (high-, moderate-, and low-impact variants) using the Ensembl-variant effect predictor (VEP, ver.103).Figure 3 Comparison of genome-wide mutations such as indels and base substitutions induced by gRNA3/Cas9 editing in TSCE5 cells under nonsynchronous and S–G2/M arrest, followed by cyclin-dependent kinase 1 inhibition (S–G2/M) conditions. Genome-wide mutations were extracted by comparing nonedited control and gRNA3/Cas9-edited cells, using Mutect2 software in GATK. The vcf files were functionally annotated using Ensembl variant effect predictor (VEP, release 103, GRCh38.p13). Annotated files were filtered to remove variants marked with the impact “modifier” and to delete redundant variants. Resulting files were manually inspected, and variants were defined as “high,” “moderate,” and “low.” High-impact variants are “stop codon gained/stopped” and “splicing donor/acceptor.” Descriptions of variant consequences follow calculated variant consequences in Ensembl VEP. Low-impact variants were synonymous or splice-region variants, the latter containing intron, noncoding, UTR, and synonymous variants (italicized). gRNA3-/Cas9–edited cells generated 859 mutations under non-syn conditions, whereas cells exhibited only five mutations under S–G2/M conditions. Result shows variants in gRNA3/Cas9-edited cells were significantly reduced under S–G2/M conditions. Similar results were obtained in I-SceI–edited cells (Fig. S5). Data are from two independent experiments. gRNA, guide RNA.

A total of 51,756 mutations (incl. modifier impact mutations that do not affect cellular functions) with redundancy were detected in gRNA3/Cas9-edited cells (versus nonedited TSCE5 cells). Among them, 859 nonredundant mutations that can affect cellular functions were observed under non-syn conditions in gRNA3/Cas9-edited cells. In contrast, under S∼G2/M conditions, only five nonredundant mutations out of 807 mutations were observed to affect cellular function (Fig. 3). During these experiments, the synchronous state of the cell cycle was checked immediately before genome editing (Fig. S2). In I-SceI–edited cells, 97,120 mutations were detected, including 1665 that can affect cellular functions under non-syn conditions. In contrast, only nine out of 1233 mutations were observed to affect cellular function under S∼G2/M conditions (Fig. S3A).

Simultaneously, our examination of naturally occurring mutations in the resting test cells showed that TSCE5 cells harbored 16 mutations that could affect their cellular functions (Fig. S4), whereas retinal pigment epithelial (RPE)-1 cells, carrying an intact TP53 gene similar to TSCE5, had 29 to 35 mutations. Therefore, the 859 mutations in gRNA3/Cas9-edited TSCE5 cells under non-syn conditions were much higher than the 16 mutations in resting TSCE5 cells. Among the genetic mutations detected in the gRNA3/Cas9-edited and I-SceI–edited cells under non-syn conditions, 113 genes, including ATM and TP53BP1, were commonly observed (Fig. S3B). Under S–G2/M conditions, genome-wide mutations caused by genome-editing tools were effectively suppressed.

Mutation signatures in gRNA3/Cas9-edited cells under two culture conditions

Cells repair naturally occurring mutations caused by oxidation, alkylation, and replication error (34, 35), such as the formation of 8-oxoguanine, O6-methylguanine and pyrimidine dimer, and oxidative deamination of cytosine. These mutations are corrected by BER, NER, and mismatch repair (MMR).

Mutations were detected in gRNA3/Cas9-edited and I-SceI–edited cells under non-syn conditions. To determine whether genome editing caused specific mutation patterns, we examined the mutation signatures of these cells. Single and doublet base substitution signatures were analyzed using SigProfiler Matrix Generator tools. Numerous single-base substitutions were generated in gRNA3/Cas9-edited and I-SceI–edited cells under non-syn conditions (Fig. 4A). Most base substitutions were C:G to A:T transversion (Fig. 4A, right panel). Ratios of transition to transversion (Ts/Tv) were 0.05 and 0.03 in gRNA3/Cas9-edited cells I-SceI–edited cells, respectively. Replication errors and DNA oxidation/alkylation could cause the accumulation of C:G to A:T transversion, such as 8-oxoguanine. Similar signatures to our result were reported in patients with mutations of POLD1 and MUTYH, defective DNA MMR/BER, and with the habit of tobacco smoking producing reactive oxygen species in COSMIC catalog of somatic mutations in cancer (ver. 96). Mutations in POLD1, which is involved in DNA replication and repair, and MUTYH, which is involved in BER, were not observed in gRNA3/Cas9-and I-SceI–edited cells (Table S1). Numerous C:G to A:T transversions were not caused by mutations in these genes. In contrast, under S–G2/M conditions, Ts/Tv was 1.09 and 0.89 in gRNA3/Cas9- and I-SceI–edited cells, respectively (Fig. 4B) and these ratios are common.Figure 4 Single-base substitution signatures generated in genome-edited cells under nonsynchronous and S–G2/M arrest followed by CDK1 inhibition (S–G2/M) conditions.A, SBS profiles of gRNA3/Cas9- and I-SceI–edited cells under non-syn conditions. Most substitutions were C:G to A:T conversions (transversion, Tv). Ratios of transition to transversion (Ts/Tv) were 0.03 to 0.05. B, SBS profiles under S–G2/M conditions. Substitutions were much lower than under asynchronous conditions. No distinct substitution patterns were observed. Ts/Tv were 1.09 and 0.89. gRNA, guide RNA.

Genome rearrangements are suppressed in cells under S∼G2/M conditions

Structural variants (SVs) of gRNA3/Cas9-edited and I-SceI–edited cells were analyzed under non-syn and S∼G2/M conditions. Under non-syn conditions, one interchromosomal translocation between chromosome (chr) 3 and chr 13 and seven intrachromosomal rearrangements were detected in gRNA3/Cas9-edited cells (Fig. 5). In contrast, under S∼G2/M conditions, only one intrachromosomal rearrangement was observed at chr 17. In addition, in I-SceI–edited cells, seven and two intrachromosomal rearrangements were detected under non-syn and S∼G2/M conditions, respectively.Figure 5 Structural variations detected in genome-edited TSCE5 cells. Circus plots showing one interchromosomal translocation between chromosome 3 (chr3) and chr13 and seven intrachromosomal rearrangements in gRNA3/Cas9-edited cells. Seven intrachromosomal rearrangements were observed in I-SceI–edited cells under non-syn conditions (left panel). In S–G2/M conditions, one and two genome rearrangements were detected in gRNA3/Cas9- and I-SceI–edited cells, respectively (right panel). Structural variants (SVs) were visualized using JBrowse2. gRNA, guide RNA.

The number of genome rearrangements was lower under S–G2/M conditions than it was under non-syn conditions in both gRNA3/Cas9-edited and I-SceI–edited cells. (Table S3). Genome rearrangements were suppressed under S–G2/M conditions. Moreover, the investigation of copy number variations (CNVs) showed no alterations or oscillations in the gRNA3/Cas9-edited cells under non-syn and S–G2/M conditions (Fig. S6).

FISH analysis reveals no chromosome aberrations in gRNA3/Cas9-edited cells

The FISH analysis to further confirm the absence of structural chromosomal rearrangements in gRNA3/Cas9-edited cells revealed no chromosomal breaks, fusions, and micronuclei under non-syn and S–G2/M conditions (Fig. 6). This result is consistent with the findings of the genome analysis (Fig. 5). This indicates that genome editing under S∼G2/M conditions is less harmful for cells.Figure 6 Chromosome aberrations in gRNA3/Cas9-edited cells. FISH analysis of gRNA3/Cas9-edited cells under non-syn and S–G2/M conditions. Chromosomal cleavage, fusion, and micronucleus formation were not observed in gRNA3/Cas9-edited TSCE5 cells. Experiment was repeated twice. Representative images are shown. The scale bar represents 10 μm. gRNA, guide RNA.

Cell cycle synchronization with CDK1 inhibition activates alternative DNA repair genes

We conducted a transcriptome analysis to identify the genes involved in the suppression of numerous genome-wide mutations. Furthermore, to elucidate the role of cell cycle synchronization and CDK1 inhibition in this process, the analysis included an additional culture condition involving Ro3306 treatment alone (Fig. 7A). Gene ontology (GO) analysis of the transcriptome data using the database for annotation, visualization, and integrated discovery revealed a reduction in the expression of genes associated with DNA replication and initiation. Gene expression profiles in gRNA3/Cas9-edited cells under S–G2/M conditions compared to nonedited control cells (Fig. S4).Figure 7 Gene profiles of cells under S–G2/M conditions or cells treated with Ro3306 alone. Transcriptome analysis was performed using two different synchronization methods: double thymidine block (d-thy)/nocodazole (noc) treatment, followed by Ro3306 treatment and Ro3306 alone. Total RNA was extracted from control cells, d-thy/noc/Ro3306-treated cells (i.e., S–G2/M conditions), and Ro3306-treated cells. Numerous genes were upregulated or downregulated in d-thy/noc/Ro3306-treated cells. Genes were extracted under conditions of log2FoldChange >1 and false discovery rate (FDR) < 0.05. A, d-thy/noc followed by Ro3306 treatment comprehensively suppressed genes related to DNA replication and cell division and activated genes for DNA repair (a). Ro3306 treatment suppressed cell division but did not block DNA replication initiation in DNA-damaged cells (b). Cell cycle synchronization suppressed genome-wide mutations. Transcriptome data were obtained two in A (a) and four in A (b) independent experiments. B, gene ontology (GO) analysis of cells under S–G2/M conditions revealed downregulation of genes associated with DNA replication and cell cycle. Expression of minichromosome maintenance 2 to 7 (Mcm2–7), cell division cycle 6 to 7 (CDC6–7), CDC45, GINS complex subunit 2 (GINS2), origin recognition complex 1 (ORC1), and downstream neighbor of Son (DONSON) genes required for Cdc45–Mcm2 to 7–GINS (CMG) complex formation and DNA replication initiation was significantly suppressed in cells under S–G2/M conditions. In contrast, zinc finger protein 384 (ZNF384), ZNF432, and dual specificity phosphatase (DUSP) were activated in these cells. ZNF384 and ZNF432 are considered to be involved in repair of ssDNA cleavage. DUSP6 modulates DNA damage response. C, GO analysis of Ro3306-treated cells. Few genes associated with DNA replication, including CDC6, MCM6, and DONSON, were downregulated. Gene expression profiles of Ro3306-treated cells were distinct from those of cells under S–G2/M conditions No significantly upregulated terms were found in GO analysis.

These genes, including minichromosome maintenance 2 to 7 (MCM2-7), MCM8-9, cell division cycle 6 (CDC6), CDC45, origin recognition complex 1, and GINS complex subunit 2, and the protein downstream neighbor of Son (DONSON), were significantly reduced in gRNA3/Cas9-edited cells under S–G2/M conditions (Fig. 7B, left table). The functions of these molecules are well documented in the literature (36, 37, 38, 39, 40, 41, 42). Conversely, the expression of genes involved in DNA repair and damage responses, such as ZNF384, ZNF432, and dual specificity phosphatase (DUSP) genes, was significantly upregulated in gRNA3/Cas9-edited cells under S–G2/M conditions (43, 44, 45, 46, 47, 48). Additionally, O-6-methylguanine methyltransferase (MGMT) was activated (log2FC = 2.58, false discovery rate = 1.5 × 10−4), and the gene expression of xeroderma pigmentosum complementation group C (XPC), which interacts with unpaired DNA bases, was significantly upregulated (log2FC = 1.83, false discovery rate = 3.4 × 10−4). These three genes are involved in BER and NER. The result suggests that ZNF and DUSP proteins enhance DNA repair capacity in gRNA3/Cas9-edited cells under S∼G2/M conditions where DNA replication is suppressed.

Cells treated with Ro3306 alone showed no changes in these genes. No significant GO terms were observed in cells treated with Ro3306 alone. Ro3306, which inhibits CDK1/cyclin B1 activity by binding to the ATP pocket of CDK1, consequently prevents mitosis (49).

Discussion

TSCE5 cells, used to investigate whether artificial DNA cleavage induces genome-wide mutations, exhibited deletions greater than 81 bp, as well as small mutations such as base substitutions within exon 5. Genome editing tools initiate DNA breaks at the target site, and while most of the DNA breaks are repaired, the breaks are then induced again. Repeated breaks and repairs can attenuate DNA repair ability, preventing the correction of naturally occurring mutations. (Fig. 8).Figure 8 TSCE5 cells were edited by gRNA3/Cas9 and selected with TFT.A, phenotype of resulting gRNA/Cas9-edited cells was derived from homozygous thymidine kinase genotype (TK−/−), containing cells with disrupted exon 5. Some cells had deletions > 81 bp at target site, whereas others had small mutation in exon 5 of thymidine kinase gene (TK). B, gRNA/Cas9-edited cells induced mutations within exon 5 of TK, visualized using Integrative Genomics Viewer (IGV). gRNA, guide RNA; TFT, trifluorothymidine.

This study provides two key insights into the behavior of genome-wide mutations during gRNA3/Cas9 editing. First, although gRNA3/Cas9 editing induces genome-wide mutations and increases the frequency of C:G to A:T transversions under standard cell culture conditions (non-syn conditions), these mutations were suppressed under S-G2/M conditions. Second, the combination of S–G2/M arrest and CDK1 inhibition using Ro3306 (S–G2/M conditions), rather than Ro3306 treatment alone, enhances the expression of emergent DNA repair genes, such as ZNF384, ZNF432, and DUSP6. To address the challenges associated with the development and accumulation of genome-wide mutations during genome editing, we analyzed the genome sequences, mutation signatures, and SVs of non-edited and genome-edited cells using gRNA3/Cas9 and I-SceI.

Cells are well known to repair over 50,000 naturally occurring mutations daily to prevent accumulation of DNA damage and ensure survival (35, 50). These naturally occurring mutations are caused by oxidation, alkylation, and replication errors, such as the formation of 8-oxoguanine, O6-methylguanine, and pyrimidine dimers and oxidative deamination of cytosine. These mutations are corrected by BER, NER, and MMR. When DNA double-stranded breaks are induced by gRNA3/Cas9, the cells correct for additional DNA lesions.

In the resting state, TSCE5 cells retain robust DNA repair capabilities, maintaining a low mutation presence (16 mutations) (Fig. S4). However, following artificial DNA double-strand cleavage by gRNA3/Cas9, these cells initiate repair processes for both the induced and naturally occurring mutations. Repeated DNA cut and repair at the gRNA3/Cas9 target site lead to ongoing cycles of cleavage and repair in primary or p53-competent cells (3, 51). Under non-syn conditions, this cycle contributes to the accumulation of genome-wide mutations, with 859 mutations observed in gRNA3/Cas9-edited cells, compared to only 16 in nonedited TSCE5 cells. However, under S-G2/M conditions, the mutation count drops dramatically to just five (Figs. 3 and S3).

Mutation signature analysis highlights a high incidence of C to A transversions under non-syn conditions, which is likely due to the build-up of oxidized and alkylated DNA bases, such as 8-oxoguanine and O-methylguanine (Fig. 4A), which are typically repaired through BER processes. Conversely, such transversions are rare under S-G2/M conditions.

The comprehensive transcriptome analysis conducted on RPE-1 cells, which have an intact TP53 gene. This is targeted at elucidating the genetic mechanisms underlying the suppression of genome-wide mutations, including determining the role of cell cycle synchronization and CDK1 inhibition in this context. To achieve this aim, we included an additional culture condition involving Ro3306 treatment alone.

This analysis reveals a significant suppression of genes critical for DNA replication initiation, such as CDC45, MCM2-7, and GINS complex subunit 2, which are essential for Cdc45–Mcm2-7–GINS complex formation, along with DONSON, origin recognition complex 1, and CDC6, which were significantly downregulated in gRNA3/Cas9-edited cells under S-G2/M conditions (Fig. 7B, left). The functions of these molecules are well-documented in the literature (36, 37, 38, 39, 40, 41, 42). No such downregulation was observed in cells exposed to Ro3306 treatment alone, except for CDC6, MCM6, and DONSON, which suggests that Ro3306 suppressed DNA replication and initiation, although the inhibition of mitosis, may not be critical (49).

In contrast, ZNF proteins are significantly upregulated under S-G2/M conditions (Fig. 7B, right). ZNF384, a member of the C2H2 family of ZNF proteins, plays a crucial role in DNA repair by binding to DNA break ends and promoting Ku70/80 assembly. The unfolding of chromatin, facilitated by PARP1, enhances the DNA-binding capacity of ZNF384 at DNA damage sites (48). ZNF432 binds to ssDNA and impedes DNA resection. These ZNF proteins are integral to both NHEJ and HR repair pathways, underscoring their significance in genome stability maintenance (47).

Furthermore, the activation of genes associated with NER and BER, such as significantly upregulated XPC and MGMT genes, suggests an enhanced capacity for repairing naturally occurring mutations under S-G2/M conditions. Additionally, proteins such as DUSP6 play a pivotal role in modulating the DNA damage response, contributing to the heightened repair capabilities observed (43, 44). These results suggest that ZNF, DUSP, MGMT, and XPC enhance the DNA repair capacity in gRNA3/Cas9-edited cells under S–G2/M conditions.

In conclusion, this study demonstrates that gRNA3/Cas9 editing not only induces genome-wide mutations but also reveals the conditional dynamics of their suppression. Under non-syn conditions, the significant increase in genome-wide mutations, particularly C:G to A:T transversions, highlights the vulnerability of genome stability to the editing processes. However, under S-G2/M conditions, where cell cycle synchronization in the S-G2/M phase and CDK1 inhibition are combined, these mutations dramatically reduce, and this is accompanied by the upregulation of DNA repair genes such as ZNF384, ZNF432, and DUSP6. These findings underscore the critical role of cell cycle control and specific DNA repair mechanisms in mitigating the unintended consequences of genome editing. Furthermore, cell cycle control with CDK1 inhibition enhances the precision and safety of CRISPR technologies in genetic research and therapeutic applications.

Experimental procedures

Cells

TSCE2, TSCE5, and TSCE122 cells were kindly provided by Masamitsu Honma and Manabu Yasui of the Division of Genetics and Mutagenesis at National Institute of Health Sciences. This study did not generate any new reagents.

Cell culture and cell cycle synchronization

TSCE5 cells, which were generated from TK6 cells by integrating the I-SceI site into intron 4 (81 bp upstream of exon 5) of TK1 via HR, were used in this study. Additionally, we designed gRNA for CRISPR/Cas9 (designated as gRNA3/Cas9) in the region between the I-SceI insertion sequence and the genomic sequence of TSCE5 cells. Briefly, the cells were maintained in RPMI 1640 medium and those with a dysfunctional TK1 gene were selected using the TK assay described in the next section. The target sites for gRNA3/Cas9 and I-SceI were inserted into the functional allele of the TK1 gene (allele A), 81 bp before exon 5, to ensure that the formation of larger deletions > 81 bp around the target sites or the occurrence of mutations within exon 5 disrupted the TK1 gene rendering it dysfunctional and knocked out as described in Figure 1, leading to the TK−/− genotype. Subsequently, the cells transformed into the TK−/− genotype survived in the presence of TFT.

To synchronize cells in the S∼G2/M phase, they were cultured with 1.5 mM of thymidine for 15 to 18 h twice with a 10-h release (double thymidine block [d-thy]) and maintained for 2 h with colcemid (0.1 μg/ml, Gibco, Thermo Fisher Scientific). Plasmids encoding I-SceI (pCBASceI, #26477) or gRNA3/Cas9 (p x 459, #48139; both from Addgene) were electroporated into the cells for genome editing. Ro3306 (5 μM, Sigma-Aldrich) was added to the S∼G2/M-synchronized cells after the electroporation, followed by incubation overnight. Cell cycle status was visualized using the cell-clock cell cycle assay kit (Biocolor). D-thy alone does not synchronize the cell cycle and, therefore, this was followed by a short colcemid treatment (because prolonged exposure is cytotoxic) to synchronize the cell cycle at the S–G2/M phase. This combination treatment is a better method for synchronizing TSCE5 cells.

TSCE2 and TSCE122 cells repaired via the HR pathway were selected in the presence of 2′-deoxycytidine and hypoxanthine-aminopterin-thymidine (cytidine and HAT). Human telomerase reverse transcriptase RPE cells were used for transcriptome analyses. RPE-1 cells were maintained in Dulbecco's modified Eagle's medium/F-12 in the presence of hygromycin B (10 μg/ml, InvivoGen). To synchronize cells in the S∼G2/M phase, they were cultured twice with 1.5 mM of thymidine for 14 to 19 h each time with a 9-h release (d-thy) and maintained for 5 h with nocodazole (noc, 50 ng/ml, Sigma-Aldrich). After genome editing, Ro3306 was added to suppress cell division and the cells were maintained for 16 h.

TK assay

The TK assay was performed using an in vitro mammalian cell gene mutation test with the TK gene (OECD validated test No. 490). The genome editing efficiency of TSCE5 is shown in Fig. S1B. TSCE5 cells were treated with cytidine and HAT for 3 days to completely suppress background TK−/− cells, electroporated with plasmid encoding gRNA3/Cas9 or I-SceI (3 μg) using the Amaxa Nucleofector 2b transfection device (program No. A-023, Lonza), and then cultured for 72 h in T-75 flasks. The cells were then seeded in 96-well plates at a density of 20,000 cells/well in the presence of TFT (2 μg/ml) and cultured for another 10 days. The number of wells without colonies (Ew) was counted to calculate the MF. In parallel, another set of electroporated cells was seeded into two 96-well plates at a density of one cell/well, and the number of wells without colonies (Ew') was calculated.

The MF was calculated as -ln (Ew/Tw)/20,000 ÷ -ln (Ewʹ/Twʹ)/1, where Ew and Tw indicate the number of wells without any colonies and the total number of wells, respectively. After TFT selection, the cells formed either small or large colonies, which indicated a potential mutation at a locus involved in cell growth or normal cell growth, respectively. Surviving TK−/− cells were selected and cultured for several days to extract genomic DNA for NGS analysis. The same experiment was performed in S–G2/M-synchronized TSCE5 cells in the presence of Ro3306, which was added after genome editing and the cells were incubated overnight. The medium was then discarded, and fresh medium was added.

Amplicon sequencing

Low-frequency mutation patterns were examined around the target site of SpCas9/gRNA3 using the amplicon sequence technique. A 10 kb region was amplified using PrimeSTAR GXL DNA polymerase (Takara Bio), and the products were purified using the NucleoSpin gel and PCR clean-up kit (Takara Bio). The purified amplicons (5–10 kb) were sequenced as 151 bp pair-end reads using the iSeq 100 system (Illumina) to obtain 12,000 coverages.

After checking the quality of the reads and trimming the reads using the FastQC (ver. 0.11.9) and Trim Galore (ver. 0.6.7) programs, respectively, the reads were mapped using the BWA-MEM (ver. 0.7.17, r1188) program. PCR duplicates were removed using the Picard software (ver. 2.26, https://broadinstitute.github.io/picard/). As a reference sequence, we used 10 kb of the GRCh38/hg38 sequence centered on the target sequence with insertion of the I-SceI site. CRISPResso2 (ver.2.0.31) was used to identify mutations at the target site of SpCas9/gRNA3 (Fig. S1B). SVs were analyzed using Manta (Illumina, ver.1.6.0, https://github.com/Illumina/manta).

Genome sequence analysis

Whole genomes of the surviving cells were analyzed in detail using NGS. The mutation ratio, size, and patterns, as well as complex genome rearrangements, were investigated in nonedited control cell and gRNA3/Cas9-and I-SceI–edited cells under the two cell culture conditions, non-syn and S–G2/M.

For the analysis, genomic DNA was extracted using the Qiagen blood and tissue kit (Qiagen), and all samples were sequenced using the Illumina Hiaseq 2500 sequencing system. The read coverage was approximately 40× and all bioinformatic analyses were conducted using a Linux operating system (Ubuntu, ver. 20.04) and the R software package (ver. 4.1.2 and later, https://www.r-project.org/). Genome sequences were analyzed and variants present only in gRNA3/Cas9- and I-SceI–edited cells were extracted by comparing the control (plasmid alone) with the genome-edited samples.

Genome sequences were analyzed using the GATK program (ver. 4.1.9.0 and 4.2.2.0) with the best practice workflow. During the analysis, reads sequenced in different lanes of the flow cells were separately handled to build unmapped and aligned binary alignment map (BAM) files, and the resulting two BAM files were merged at the MarkDuplicates step in the Picard program (ver. 2.25). Paired-end reads (150 bp × 2) were aligned to the human reference genome (GRCh38/hg38 for GATK, Google Cloud Bucket) using the BWA-MEM program (ver. 0.7.17, r1188) with the M option. PCR duplicates were marked using MarkDuplicates after the aligned and unaligned BAM files were merged. MergeBamAlignment was run using the options “MAX_INSERTIONS_OR_DELETIONS −1,” “-PRIMARY_ALIGNMENT_STRATEGY MostDistant,” and “CLIP_ADAPTOR false.” The prepared clean BAM files were analyzed using GATK/Mutect2, followed by FilterMutectCalls in GATK to detect somatic variants, such as small substitutions, deletions, and insertions. Mutect2 was run with the option “af-of-alleles-not-in-resource −1.0,” where the tool dynamically adjusts the parameters for an appropriate mode. The af-of-alleles-not-in-resource option was tested at 0.000025 (previously recommended) for comparison; however, both results were the same. The impacts of variants detected were evaluated by the Ensembl-VEP (release 103).

Mutational signatures, including all types of somatic mutations, were analyzed using the SigProfilerMatrixGenerator (COSMIC, Wellcome Sanger Institute, ver.3.2).

SVs and CNVs were analyzed using the following pipelines: SvABA (Broad Institute, ver.1.1) (52) for SVs and ModelSegments CNV Workflow in GATK for CNVs. The software for SV detection was run in somatic modes; for example, the option of -a somatic_run_mode2 in SvABA. The filter-passed SVs were subjected to further analyses. The extraction of passed variants was conducted using SnpEff and SnpSift (ver.5.1). For CNVs, read counts were collected and denoised to generate ModeldSegments file. The CNV results were plotted using the PlotDenoisedCopyRatios tool. Delly (EMBL, ver.0.8.9) and Manta (Illumina, ver. 1.6) for SV analysis were used for comparison.

Circos plots were constructed using the R package Biocircos (ver.0.3.4) (53). R scripts and packages were executed using R studio (Build 443). R was also used for general data handling, such as removing redundant data from Ensembl-VEP and extracting specific characters from the DataFrame, including the extraction of G from atG and Gca.

Mutational signatures, including all types of somatic mutations, were analyzed using the SigProfilerMatrixGenerator (COSMIC, Wellcome Sanger Institute, ver.3.2).

Long-read sequences were analyzed using the PromethION (Oxford Nanopore). The Guppy program(ver.4.4.1) was used to convert fast5 filles to fastq files, which were trimmed with the options of -q 9 and headcrop 75 using Nanofilt (ver.2.8.0). The trimmed reads were aligned to the reference using Minimap2 (ver.2.24) (54). Sniffles (ver.2.0.7) (55) with the options of --minsupport 5, --minsvlen 50, –nongermline were used for SV calls. Somatic variants were obtained by subtracting variants in the control sample from genome-edited samples using bcftools (ver.1.12). The NGS reads were visualized using the Integrative Genomics Viewer (ver.2.13.2) (56) and JBrowse2 (57) were used.

Transcriptome analysis

Transcriptome analysis was performed for RPE-1 cells under two different synchronization conditions: d-thy/noc treatment followed by Ro3306 treatment or Ro3306 alone. The data were compared with those from non-syn control cells. Total RNA was extracted from control cells and d-thy/noc/Ro-treated cells (i.e., S∼G2/M conditions) or Ro-treated cells using the RNeasy mini kit (QIAGEN). After a quality check of RNAs (RIN>7) using the Bioanalyzer (Agilent), complementary DNA samples were prepared for RNA-seq using the SMART-Seq v4 Ultra Low Input RNA kit. RNA-seq was performed using the NovaSeq 6000 sequencing system (Illumina). The sequence reads were mapped to GRCh38 using the DRAGEN Bio-IT platform (ver.3.6.3). Expression levels are shown as transcripts per million, and genes with more than a 2-fold change between the control and genome-edited groups were extracted. The experimental conditions were the same as those described above. Functional annotation was performed using the database for annotation, visualization, and integrated discovery tool (LHRI) (58). PANTHER was also used for GO analysis to compare the two results (59).

FISH analysis

FISH was performed according to the Agilent FISH protocol. Briefly, cells were harvested, gently mixed with 0.075 M potassium chloride, fixed with cold methanol–acetic acid (3:1) solution, and stored overnight at −20 °C. The following day, the cells were resuspended, dropped onto glass slides soaked in cooled ethanol, and dried overnight. The slides were then fixed with 4% formaldehyde; washed twice; sequentially dehydrated in 70%, 85%, and 100% ethanol for 1 min each; and then dried.

The slides were subsequently coverslipped by placing then on a cover glass with a hybridization mix containing the TelC-Cy3 telomere probe (Cat# F1002, PANAGENE) and then turning the slides over. The glass slides were placed on a hot plate, heated at 80 °C for 5 min, and then placed on a paper towel for 1 h. The slides were then incubated in FISH wash buffer (Agilent), heated at 73 °C for several minutes, dried with ethanol, and mounted using Pro Long Gold (Invitrogen, Thermo Fisher Scientific) with 4′,6-diamidino-2-phenylindole.

Data availability

Our dataset used for analysis is available at the DNA Data Bank of Japan (DDBJ, PRJDB18589). We can also provide bam formatted files upon request. Reference human genome sequence was obtained from GATK resource bundle at https://console.cloud.google.com/storage/browser/genomics-public-data/resources/broad/hg38/v0/Homo_sapience_assembly38.fasta.

Conflicts of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Supporting information

Supplemental information

Table S1

Table S2

Table S3

Table S4

Acknowledgments

We thank Masamitsu Honma and Manabu Yasui (Division of Genetics and Mutagenesis, National Institute of Health Sciences for giving us TSCE cells and helpful suggestion. We also thank all members of our lab (NIHS, Division of Biochemistry) for technical supports and helpful discussion during this study.

Author contributions

N. F. and K. K. conceptualization; N. F., K. S., and K. K. methodology; N. F., K. Soga, C. T., J. N., K. Sakata, and R. K. investigation; K. Soga, C. T., J. N., K. Sakata, R. K., S. Y., N. S., and K. K. formal analysis; K. Soga, J. N., S. Y., N. S., and K. K. visualization; K. K. supervision; K. Soga, J. N., S. Y., N. S., and K. K. writing–original draft; K. K. funding acquisition; K. K. writing–review and editing; K. K. project administration.

Funding and additional information

This work was supported in part by 10.13039/501100003478 MHLW (Grant Number 21KA1002 ) and the 10.13039/501100001691 Japan Society for the Promotion of Science and the 10.13039/501100001700 Ministry of Education, Culture, Sports, Science and Technology (KAKENHI Grant Number JP21H02469 ).
==== Refs
References

1 Jinek M. Chylinski K. Fonfara I. Hauer M. Doudna J.A. Charpentier E. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity Science 337 2012 816 821 22745249
2 Cong L. Ran F.A. Cox D. Lin S. Barretto R. Habib N. Multiplex genome engineering using CRISPR/Cas systems Science 339 2013 819 823 23287718
3 Sternberg S.H. Redding S. Jinek M. Greene E.C. Doudna J.A. DNA interrogation by the CRISPR RNA-guided endonuclease Cas9 Nature 507 2014 62 67 24476820
4 Singh D. Sternberg S.H. Fei J. Doudna J.A. Ha T. Real-time observation of DNA recognition and rejection by the RNA-guided endonuclease Cas9 Nat. Commun. 7 2016 12778
5 Ivanov I.E. Wright A.V. Cofsky J.C. Aris K.D.P. Doudna J.A. Bryant Z. Cas9 interrogates DNA in discrete steps modulated by mismatches and supercoiling Proc. Natl. Acad. Sci. U. S. A. 117 2020 5853 5860 32123105
6 Kleinstiver B.P. Pattanayak V. Prew M.S. Tsai S.Q. Nguyen N.T. Zheng Z. High-fidelity CRISPR-Cas9 nucleases with no detectable genome-wide off-target effects Nature 529 2016 490 495 26735016
7 Vakulskas C.A. Dever D.P. Rettig G.R. Turk R. Jacobi A.M. Collingwood M.A. A high-fidelity Cas9 mutant delivered as a ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells Nat. Med. 24 2018 1216 1224 30082871
8 Schmid-Burgk J.L. Gao L. Li D. Gardner Z. Strecker J. Lash B. Highly parallel profiling of Cas9 variant specificity Mol. Cell 78 2020 794 800.e798 32187529
9 Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system Cell 163 2015 759 771 26422227
10 Kim D. Kim J. Hur J.K. Been K.W. Yoon S.H. Kim J.S. Genome-wide analysis reveals specificities of Cpf1 endonucleases in human cells Nat. Biotechnol. 34 2016 863 868 27272384
11 Kleinstiver B.P. Tsai S.Q. Prew M.S. Nguyen N.T. Welch M.M. Lopez J.M. Genome-wide specificities of CRISPR-Cas Cpf1 nucleases in human cells Nat. Biotechnol. 34 2016 869 874 27347757
12 Kleinstiver B.P. Sousa A.A. Walton R.T. Tak Y.E. Hsu J.Y. Clement K. Engineered CRISPR-Cas12a variants with increased activities and improved targeting ranges for gene, epigenetic and base editing Nat. Biotechnol. 37 2019 276 282 30742127
13 Kim N. Kim H.K. Lee S. Seo J.H. Choi J.W. Park J. Prediction of the sequence-specific cleavage activity of Cas9 variants Nat. Biotechnol. 38 2020 1328 1336 32514125
14 Kim D. Bae S. Park J. Kim E. Kim S. Yu H.R. Digenome-seq: genome-wide profiling of CRISPR-Cas9 off-target effects in human cells Nat. Methods 12 2015 237 243 25664545
15 Tsai S.Q. Zheng Z. Nguyen N.T. Liebers M. Topkar V.V. Thapar V. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases Nat. Biotechnol. 33 2015 187 197 25513782
16 Wang X. Wang Y. Wu X. Wang J. Wang Y. Qiu Z. Unbiased detection of off-target cleavage by CRISPR-Cas9 and TALENs using integrase-defective lentiviral vectors Nat. Biotechnol. 33 2015 175 178 25599175
17 Cameron P. Fuller C.K. Donohoue P.D. Jones B.N. Thompson M.S. Carter M.M. Mapping the genomic landscape of CRISPR-Cas9 cleavage Nat. Methods 14 2017 600 606 28459459
18 Wienert B. Wyman S.K. Richardson C.D. Yeh C.D. Akcakaya P. Porritt M.J. Unbiased detection of CRISPR off-targets in vivo using DISCOVER-Seq Science 364 2019 286 289 31000663
19 Atkins A. Chung C.H. Allen A.G. Dampier W. Gurrola T.E. Sariyer I.K. Off-target analysis in gene editing and applications for clinical translation of CRISPR/Cas9 in HIV-1 therapy Front. Genome Ed. 3 2021 673022
20 van Overbeek M. Capurso D. Carter M.M. Thompson M.S. Frias E. Russ C. DNA repair profiling reveals nonrandom outcomes at Cas9-mediated breaks Mol. Cell 63 2016 633 646 27499295
21 Taheri-Ghahfarokhi A. Taylor B.J.M. Nitsch R. Lundin A. Cavallo A.L. Madeyski-Bengtson K. Decoding non-random mutational signatures at Cas9 targeted sites Nucleic Acids Res. 46 2018 8417 8434 30032200
22 Liu M. Zhang W. Xin C. Yin J. Shang Y. Ai C. Global detection of DNA repair outcomes induced by CRISPR-Cas9 Nucleic Acids Res. 49 2021 8732 8742 34365511
23 Stinson B.M. Moreno A.T. Walter J.C. Loparo J.J. A mechanism to minimize errors during non-homologous end joining Mol. Cell 77 2020 1080 1091.e1088 31862156
24 Shibata A. Conrad S. Birraux J. Geuting V. Barton O. Ismail A. Factors determining DNA double-strand break repair pathway choice in G2 phase EMBO J. 30 2011 1079 1092 21317870
25 Kosicki M. Allen F. Steward F. Tomberg K. Pan Y. Bradley A. Cas9-induced large deletions and small indels are controlled in a convergent fashion Nat. Commun. 13 2022 3422 35701408
26 Yu W. Lescale C. Babin L. Bedora-Faure M. Lenden-Hasse H. Baron L. Repair of G1 induced DNA double-strand breaks in S-G2/M by alternative NHEJ Nat. Commun. 11 2020 5239 33067475
27 OECD Test No. 490: in vitro mammalian cell gene mutation tests using the thymidine kinase gene OECD Guidelines for the Testing of Chemicals, Section 4 2016 OECD Paris, France 10.1787/20745788
28 Honma M. Izumi M. Sakuraba M. Tadokoro S. Sakamoto H. Wang W. Deletion, rearrangement, and gene conversion; genetic consequences of chromosomal double-strand breaks in human cells Environ. Mol. Mutagen 42 2003 288 298 14673874
29 Li Y. Lian D. Wang J. Zhao Y. Li Y. Liu G. MDM2 antagonists promote CRISPR/Cas9-mediated precise genome editing in sheep primary cells Mol. Ther. Nucleic Acids 31 2023 309 323 36726409
30 Smirnikhina S.A. Zaynitdinova M.I. Sergeeva V.A. Lavrov A.V. Improving homology-directed repair in genome editing experiments by influencing the cell cycle Int. J. Mol. Sci. 23 2022 5992 35682671
31 Matsumoto D. Tamamura H. Nomura W. A cell cycle-dependent CRISPR-Cas9 activation system based on an anti-CRISPR protein shows improved genome editing accuracy Commun. Biol. 3 2020 601 33097793
32 Yang D. Scavuzzo M.A. Chmielowiec J. Sharp R. Bajic A. Borowiak M. Enrichment of G2/M cell cycle phase in human pluripotent stem cells enhances HDR-mediated gene repair with customizable endonucleases Sci. Rep. 6 2016 21264
33 Lin S. Staahl B.T. Alla R.K. Doudna J.A. Enhanced homology-directed human genome engineering by controlled timing of CRISPR/Cas9 delivery Elife 3 2014 e04766
34 Atamna H. Cheung I. Ames B.N. A method for detecting abasic sites in living cells: age-dependent changes in base excision repair Proc. Natl. Acad. Sci. U. S. A. 97 2000 686 691 10639140
35 Barnes D.E. Lindahl T. Repair and genetic consequences of endogenous DNA base damage in mammalian cells Annu. Rev. Genet. 38 2004 445 476 15568983
36 Reynolds J.J. Bicknell L.S. Carroll P. Higgs M.R. Shaheen R. Murray J.E. Mutations in DONSON disrupt replication fork stability and cause microcephalic dwarfism Nat. Genet. 49 2017 537 549 28191891
37 Hashimoto Y. Sadano K. Miyata N. Ito H. Tanaka H. Novel role of DONSON in CMG helicase assembly during vertebrate DNA replication initiation EMBO J. 42 2023 e114131
38 Lim Y. Tamayo-Orrego L. Schmid E. Tarnauskaite Z. Kochenova O.V. Gruar R. In silico protein interaction screening uncovers DONSON's role in replication initiation Science 381 2023 eadi3448
39 Yuan Z. Riera A. Bai L. Sun J. Nandi S. Spanos C. Structural basis of Mcm2-7 replicative helicase loading by ORC-Cdc6 and Cdt1 Nat. Struct. Mol. Biol. 24 2017 316 324 28191893
40 Powell S.K. MacAlpine H.K. Prinz J.A. Li Y. Belsky J.A. MacAlpine D.M. Dynamic loading and redistribution of the Mcm2-7 helicase complex through the cell cycle EMBO J. 34 2015 531 543 25555795
41 Helderman N.C. Terlouw D. Bonjoch L. Golubicki M. Antelo M. Morreau H. Molecular functions of MCM8 and MCM9 and their associated pathologies iScience 26 2023 106737
42 Nishimura K. Ishiai M. Horikawa K. Fukagawa T. Takata M. Takisawa H. Mcm8 and Mcm9 form a complex that functions in homologous recombination repair induced by DNA interstrand crosslinks Mol. Cell 47 2012 511 522 22771115
43 Bagnyukova T.V. Restifo D. Beeharry N. Gabitova L. Li T. Serebriiskii I.G. DUSP6 regulates drug sensitivity by modulating DNA damage response Br. J. Cancer 109 2013 1063 1071 23839489
44 Robitaille A.C. Caron E. Zucchini N. Mukawera E. Adam D. Mariani M.K. DUSP1 regulates apoptosis and cell migration, but not the JIP1-protected cytokine response, during respiratory syncytial virus and sendai virus infection Sci. Rep. 7 2017 17388
45 Gulve N. Su C. Deng Z. Soldan S.S. Vladimirova O. Wickramasinghe J. DAXX-ATRX regulation of p53 chromatin binding and DNA damage response Nat. Commun. 13 2022 5033 36028493
46 Singh J.K. van Attikum H. DNA double-strand break repair: putting zinc fingers on the sore spot Semin. Cell Dev. Biol. 113 2021 65 74 32962915
47 O'Sullivan J. Kothari C. Caron M.C. Gagne J.P. Jin Z. Nonfoux L. ZNF432 stimulates PARylation and inhibits DNA resection to balance PARPi sensitivity and resistance Nucleic Acids Res. 51 2023 11056 11079 37823600
48 Singh J.K. Smith R. Rother M.B. de Groot A.J.L. Wiegant W.W. Vreeken K. Zinc finger protein ZNF384 is an adaptor of Ku to DNA during classical non-homologous end-joining Nat. Commun. 12 2021 6560 34772923
49 Vassilev L.T. Tovar C. Chen S. Knezevic D. Zhao X. Sun H. Selective small-molecule inhibitor reveals critical mitotic functions of human CDK1 Proc. Natl. Acad. Sci. U. S. A. 103 2006 10660 10665 16818887
50 Chen J. Miller B.F. Furano A.V. Repair of naturally occurring mismatches can induce mutations in flanking DNA Elife 3 2014 e02001
51 Raper A.T. Stephenson A.A. Suo Z. Functional insights revealed by the kinetic mechanism of CRISPR/Cas9 J. Am. Chem. Soc. 140 2018 2971 2984 29442507
52 Wala J.A. Bandopadhayay P. Greenwald N.F. O'Rourke R. Sharpe T. Stewart C. SvABA: genome-wide detection of structural variants and indels by local assembly Genome Res. 28 2018 581 591 29535149
53 Cui Y. Chen X. Luo H. Fan Z. Luo J. He S. BioCircos.js: an interactive Circos JavaScript library for biological data visualization on web applications Bioinformatics 32 2016 1740 1742 26819473
54 Li H. Minimap2: pairwise alignment for nucleotide sequences Bioinformatics 34 2018 3094 3100 29750242
55 Sedlazeck F.J. Rescheneder P. Smolka M. Fang H. Nattestad M. von Haeseler A. Accurate detection of complex structural variations using single-molecule sequencing Nat. Methods 15 2018 461 468 29713083
56 Thorvaldsdottir H. Robinson J.T. Mesirov J.P. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration Brief Bioinform. 14 2013 178 192 22517427
57 Diesh C. Stevens G.J. Xie P. De Jesus Martinez T. Hershberg E.A. Leung A. JBrowse 2: a modular genome browser with views of synteny and structural variation Genome Biol. 24 2023 74 37069644
58 Sherman B.T. Hao M. Qiu J. Jiao X. Baseler M.W. Lane H.C. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update) Nucleic Acids Res. 50 2022 W216 W221 35325185
59 Thomas P.D. Ebert D. Muruganujan A. Mushayahama T. Albou L.P. Mi H. PANTHER: making genome-scale phylogenetics accessible to all Protein Sci. 31 2022 8 22 34717010
