Advertisement

CRISPR–Cas9 gene editing induced complex on-target outcomes in human cells

  • Wei Wen
    Affiliations
    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Haihe Laboratory of Cell Ecosystem, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Search for articles by this author
  • Xiao-Bing Zhang
    Correspondence
    Corresponding author: Xiao-Bing Zhang, National Clinical Research Center for Blood Diseases, Haihe Laboratory of Cell Ecosystem, State Key Laboratory of Experimental Hematology, Tianjin 300020, China
    Affiliations
    State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Haihe Laboratory of Cell Ecosystem, Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
    Search for articles by this author
Open AccessPublished:March 15, 2022DOI:https://doi.org/10.1016/j.exphem.2022.03.002

      Highlights

      • Unwanted on-target mutation occurs after CRISPR–Cas9 cleavage.
      • Assessment of comprehensive on-target outcomes is necessary.
      • Clinical genomic engineering requires quality controls to address safety concerns.
      CRISPR–Cas9 is a powerful tool for editing the genome and holds great promise for gene therapy applications. Initial concerns of gene engineering focus on off-target effects. However, in addition to short indel mutations (often <50 bp), an increasing number of studies have revealed complex on-target results after double-strand break repair by CRISPR–Cas9, such as large deletions, gene rearrangement, and loss of heterozygosity. These unintended mutations are potential safety concerns in clinical gene editing. Here, in this review, we summarize the significant findings of CRISPR–Cas9-induced on-target deleterious outcomes and discuss putative ways to achieve safe gene therapy.

      CRISPR–CAS9-MEDIATED GENE EDITING AND DNA REPAIR MACHINERY

      Overview of CRISPR–Cas9

      Clustered, regularly interspaced, short palindromic repeats (CRISPR)–Cas9 has been a promising tool for gene engineering, for example, in correcting disease-associated mutant alleles in somatic or stem cells [
      • Lander ES.
      The heroes of CRISPR.
      ]. Cas9 is a single endonuclease evolved in bacteria and archaea to function as a natural adaptive immune system [
      • Deltcheva E
      • Chylinski K
      • Sharma CM
      • et al.
      CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III.
      ,
      • Gasiunas G
      • Barrangou R
      • Horvath P
      • Siksnys V.
      Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria.
      ]. Cas9 programmed with crRNA (CRISPR RNA) and tracrRNA (trans-activating crRNA) (Cas9–sgRNA ribonucleoprotein complex) has HNH and RuvC nuclease domains to cleave target DNA, generating two blunt ends of double-strand breaks (DSBs), usually 3 bp upstream of a protospacer adjacent motif (PAM, NGG for SpCas9 from Streptococcus pyogenes) sequence [
      • Gasiunas G
      • Barrangou R
      • Horvath P
      • Siksnys V.
      Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria.
      ,
      • Jinek M
      • Chylinski K
      • Fonfara I
      • Hauer M
      • Doudna JA
      • Charpentier E.
      A programmable dual-RNA–guided DNA endonuclease in adaptive bacterial immunity.
      ]. Two elements are important for the cleavage by Cas9: (1) the requirement for a guide RNA that base pairs with the target sequence; (2) the necessity of the PAM site for target identification. Cas9 nickase is an enzyme that is inactivated in either nuclease domain. Inactivation of both the HNH and RuvC domains can generate catalytically dead Cas9 (dCas9), which maintains the capacity to bind to the cognate DNA without cutting it [
      • Nishimasu H
      • Ran FA
      • Hsu PD
      • et al.
      Crystal structure of Cas9 in complex with guide RNA and target DNA.
      ,
      • Cong L
      • Ran FA
      • Cox D
      • et al.
      Multiplex genome engineering using CRISPR/Cas systems.
      ]. Nickases are engineered and used for applications such as base editors (BEs) and prime editors (PEs) [
      • Komor AC
      • Kim YB
      • Packer MS
      • Zuris JA
      • Liu DR.
      Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage.
      ,
      • Gaudelli NM
      • Komor AC
      • Rees HA
      • et al.
      Programmable base editing of A•T to G•C in genomic DNA without DNA cleavage.
      ,
      • Rees HA
      • Liu DR.
      Base editing: precision chemistry on the genome and transcriptome of living cells.
      ,
      • Anzalone AV
      • Randolph PB
      • Davis JR
      • et al.
      Search-and-replace genome editing without double-strand breaks or donor DNA.
      ]. Both BEs and PEs can edit DNA without generating DSBs. dCas9 can be used for transcriptional regulation and epigenetic modifications [
      • Shalem O
      • Sanjana NE
      • Zhang F.
      High-throughput functional genomics using CRISPR–Cas9.
      ,
      • Thakore PI
      • Black JB
      • Hilton IB
      • Gersbach CA.
      Editing the epigenome: technologies for programmable transcription and epigenetic modulation.
      ,
      • Pickar-Oliver A
      • Gersbach CA.
      The next generation of CRISPR–Cas technologies and applications.
      ]. CRISPR–Cas9 cleavage of one target site often induces nucleotide insertions and deletions. Simultaneous editing of two sites has been used to generate inversion of a DNA segment or chromosomal translocation [
      • Shou J
      • Li J
      • Liu Y
      • Wu Q.
      Precise and predictable CRISPR chromosomal rearrangements reveal principles of Cas9-mediated nucleotide insertion.
      ]. DSB-Induced repair outcomes with similar sequences can result from distinct mechanisms [
      • Hussmann JA
      • Ling J
      • Ravisankar P
      • et al.
      Mapping the genetic landscape of DNA double-strand break repair.
      ]. These processes are regulated mainly by several types of competing DNA repair machinery [
      • Nambiar TS
      • Baudrier L
      • Billon P
      • Ciccia A.
      CRISPR-based genome editing through the lens of DNA repair.
      ], including nonhomologous end joining (NHEJ), microhomology-mediated end joining (MMEJ), and homology-directed repair (HDR).

      Nonhomologous End Joining

      DSBs, one of the most hazardous forms of DNA damage, are a potent source of genome instability. Thus, to sustain an intact genome, cells with DSBs require rapid ligation of broken DNA ends and repair the break as quickly as possible. This is largely accomplished by a pathway termed canonical nonhomologous end joining (c-NHEJ), which is a rapid yet often erroneous repair process. During this process, key NHEJ molecules, such as Ku70, Ku80, and DNA ligase IV, are recruited to the ends of breaks, leading to nonmutated repair or producing mutated alleles with short insertion or deletions (indels) [
      • Scully R
      • Panday A
      • Elango R
      • Willis NA.
      DNA double-strand break repair-pathway choice in somatic mammalian cells.
      ]. When perfect end joining occurs, the reconstituted Cas9–sgRNA cognate sequence can be recut in the persistent presence of Cas9–sgRNA. However, end joining can also result in an insertion or deletion mutant outcome, preventing subsequent recognition and recutting by the nuclease [
      • Brinkman EK
      • Chen T
      • de Haas M
      • Holland HA
      • Akhtar W
      • van Steensel B.
      Kinetics and fidelity of the repair of Cas9-induced double-strand DNA breaks.
      ].
      Furthermore, when the sgRNA guides Cas9 to target an open reading frame (ORF), the indel outcome may generate a frameshift mutation that abrogates protein function [
      • Wang T
      • Wei JJ
      • Sabatini DM
      • Lander ES.
      Genetic screens in human cells using the CRISPR-Cas9 system.
      ,
      • Shalem O
      • Sanjana NE
      • Hartenian E
      • et al.
      Genome-scale CRISPR-Cas9 knockout screening in human cells.
      ] and create a premature stop codon that may trigger nonsense-mediated mRNA decay (NMD) [
      • Brogna S
      • Wen J.
      Nonsense-mediated mRNA decay (NMD) mechanisms.
      ,
      • Popp MW
      • Maquat LE.
      Leveraging rules of nonsense-mediated mRNA decay for genome engineering and personalized medicine.
      ]. Recently, multiple reports have shown that editing outcomes are nonrandom [
      • van Overbeek M
      • Capurso D
      • Carter MM
      • et al.
      DNA repair profiling reveals nonrandom outcomes at Cas9-mediated breaks.
      ] and can be predicted [
      • Allen F
      • Crepaldi L
      • Alsinet C
      • et al.
      Predicting the mutations generated by repair of Cas9-induced double-strand breaks.
      ]. The mechanism mainly relies on target sequences and is less affected by cell types. Particularly in the position where the −4 bp upstream PAM sequence is a T (thymine), +T editing is the most frequent outcome of SpCas9 [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ]. As such, most insertions result from NHEJ repair after Cas9 editing. Our unpublished work shows that this is a unique feature of SpCas9, and other Cas9 orthologs may behave differently. Due to its efficacy and predominant nature in bridging blunt ends, this end-joining machinery has been exploited to insert a long sequence using AAV [
      • Zhang JP
      • Cheng XX
      • Zhao M
      • et al.
      Curing hemophilia A by NHEJ-mediated ectopic F8 insertion in the mouse.
      ] as well as a double-strand oligonucleotide (dsODN) [
      • Auer TO
      • Duroure K
      • De Cian A
      • Concordet JP
      • Del Bene F.
      Highly efficient CRISPR/Cas9-mediated knock-in in zebrafish by homology-independent DNA repair.
      ,
      • Suzuki K
      • Tsunekawa Y
      • Hernandez-Benitez R
      • et al.
      In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration.
      ], at the target site, as in the assessment of off-targets using the GUIDE-seq assay [
      • Tsai SQ
      • Zheng Z
      • Nguyen NT
      • et al.
      GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases.
      ].

      Microhomology-Mediated End Joining

      The default engagement of c-NHEJ can be disrupted by DNA end resection. When resection occurs at the breaks, the Mre11–Rad50–Nbs1 (MRN) complex [
      • Sfeir A
      • Symington LS.
      Microhomology-mediated end joining: a back-up survival mechanism or dedicated pathway?.
      ] may be recruited, which hampers repair by the NHEJ pathway. In such scenario, microhomology-mediated end joining (MMEJ), also called alternative end joining (alt-EJ), is activated. MMEJ is considered to use stretches of microhomology of 2–25 bp on each side of the DSBs to realign the broken ends [
      • Sfeir A
      • Symington LS.
      Microhomology-mediated end joining: a back-up survival mechanism or dedicated pathway?.
      ]. However, we found that a single C or G at both ends can also mediate significant levels of MMEJ editing [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ]. The microhomologies also define the boundaries for DNA segment rejoining, leading to deletion of one microhomology and the regions between the two microhomologies. The MMEJ occurrence depends on the proximity to the break ends, and the length and G/C proportion of the microhomology, which determine the binding energy. The longer the mismatched nucleotides between the two homologies, the lower is the efficiency of MMEJ, as it will need to recruit flap endonucleases to remove DNA flaps. Therefore, a “strong” microhomology arm can generate a highly efficient MMEJ-mediated precise deletion [
      • Iyer S
      • Suresh S
      • Guo D
      • et al.
      Precise therapeutic gene correction by a simple nuclease-induced double-stranded break.
      ,
      • Wang L
      • Li L
      • Ma Y
      • et al.
      Reactivation of γ-globin expression through Cas9 or base editor to treat β-hemoglobinopathies.
      ]. This phenomenon can be exploited to bring about a predictable and ideal editing outcome.

      Homology-Directed Repair

      To obtain more precise results, either single-nucleotide changes or large gene cassette insertion/deletion at specific sites, one needs to harness the HDR pathway by providing a donor template [
      • Porteus MH.
      Towards a new era in medicine: therapeutic genome editing.
      ,
      • Dever DP
      • Porteus MH.
      The changing landscape of gene editing in hematopoietic stem cells: a step towards Cas9 clinical translation.
      ]. An HDR donor contains an insertion sequence flanked by right and left homology arms (HAs), whose sequences are preferably identical to 300–1000 bp surrounding the DSB site. HDRs can make small genetic edits [
      • Chu VT
      • Weber T
      • Wefers B
      • et al.
      Increasing the efficiency of homology-directed repair for CRISPR–Cas9-induced precise gene editing in mammalian cells.
      ] and create precise insertion of long fragments [
      • Chu VT
      • Weber T
      • Wefers B
      • et al.
      Increasing the efficiency of homology-directed repair for CRISPR–Cas9-induced precise gene editing in mammalian cells.
      ], such as a gene of interest or a fluorescent reporter.
      Although end-joining processes (including NHEJ and MMEJ) are efficient in most cell types regardless of the cell cycle, potent HDR editing occurs predominantly in the S/G2 phases of the cell cycle [
      • Takata M
      • Sasaki MS
      • Sonoda E
      • et al.
      Homologous recombination and non-homologous end-joining pathways of DNA double-strand break repair have overlapping roles in the maintenance of chromosomal integrity in vertebrate cells.
      ]. It has a relatively lower efficiency than NHEJ. However, HDR-mediated gene engineering has the advantage of enabling a large fragment DNA knockin with donor templates delivered in the form of plasmids [
      • Li XL
      • Li GH
      • Fu J
      • et al.
      Highly efficient genome editing via CRISPR-Cas9 in human pluripotent stem cells is achieved by transient BCL-XL overexpression.
      ,
      • Zhang JP
      • Li XL
      • Li GH
      • et al.
      Efficient precise knockin with a double cut HDR donor after CRISPR/Cas9-mediated double-stranded DNA cleavage.
      ,
      • Wen W
      • Cheng X
      • Fu Y
      • et al.
      High-level precise knockin of iPSCs by simultaneous reprogramming and genome editing of human peripheral blood mononuclear cells.
      ], adeno-associated virus (AAV) vectors [
      • Kuo CY
      • Long JD
      • Campo-Fernandez B
      • et al.
      Site-specific gene editing of human hematopoietic stem cells for X-linked hyper-IgM syndrome.
      ,
      • Vakulskas CA
      • Dever DP
      • Rettig GR
      • et al.
      A high-fidelity Cas9 mutant delivered as a ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells.
      ,
      • Pavel-Dinu M
      • Wiebking V
      • Dejene BT
      • et al.
      Gene correction for SCID-X1 in long-term hematopoietic stem cells.
      ,
      • Ferrari S
      • Jacob A
      • Beretta S
      • et al.
      Efficient gene editing of human long-term hematopoietic stem cells validated by clonal tracking.
      ], or long ssDNA HDR donors [
      • Quadros RM
      • Miura H
      • Harms DW
      • et al.
      Easi-CRISPR: a robust method for one-step generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR ribonucleoproteins.
      ].
      To promote HDR-mediated gene editing efficiency, a “double-cut” donor flanked by sgRNA–PAM sequences with HA can increase the HDR efficiency up to 10-fold [
      • Zhang JP
      • Li XL
      • Li GH
      • et al.
      Efficient precise knockin with a double cut HDR donor after CRISPR/Cas9-mediated double-stranded DNA cleavage.
      ]. This design synchronizes the availability of both DSBs at the genome target and the linearized donor templates. In human induced pluripotent stem cells (iPSCs), a 20%–30% HDR-mediated knockin can be obtained using double-cut donors with 300- or 600-bp-long HAs [
      • Zhang JP
      • Li XL
      • Li GH
      • et al.
      Efficient precise knockin with a double cut HDR donor after CRISPR/Cas9-mediated double-stranded DNA cleavage.
      ]. In K562 cells, a myelogenous leukemia cell line, the use of the double-cut HDR donor plasmid results in >50% editing efficacy (unpublished). With the development strategies of HDR-mediated knockin and blood cell reprogramming, blood cells can be reprogrammed into iPSCs and edited simultaneously at high efficiency in one step [
      • Wen W
      • Cheng X
      • Fu Y
      • et al.
      High-level precise knockin of iPSCs by simultaneous reprogramming and genome editing of human peripheral blood mononuclear cells.
      ].

      Dynamics and Competition of the Three Major Repair Pathways

      To repair dsDNA damage induced by CRISPR–Cas9, these three major DNA repair pathways play a complementary role [
      • Tatiossian KJ
      • Clark RDE
      • Huang C
      • Thornton ME
      • Grubbs BH
      • Cannon PM.
      Rational selection of CRISPR-Cas9 guide RNAs for homology-directed genome editing.
      ,
      • Roidos P
      • Sungalee S
      • Benfatto S
      • et al.
      A scalable CRISPR/Cas9-based fluorescent reporter assay to study DNA double-strand break repair choice.
      ]. NHEJ is a rapid and fast pathway, often leading to a +A/T editing outcome at the cutting site [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ]. On the other hand, HDR and MMEJ are activated more slowly than NHEJ [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ], mainly because HDR and MMEJ entail an Mre11-dependent DSB end resection process, yet NHEJ takes effect by directly joining two “clean” ends [
      • Truong LN
      • Li Y
      • Shi LZ
      • et al.
      Microhomology-mediated end joining and homologous recombination share the initial end resection step to repair DNA double-strand breaks in mammalian cells.
      ]. Of interest, HDR outcompetes MMEJ [
      • Truong LN
      • Li Y
      • Shi LZ
      • et al.
      Microhomology-mediated end joining and homologous recombination share the initial end resection step to repair DNA double-strand breaks in mammalian cells.
      ], likely because the presence of long homology in HDR stabilizes the annealing of donor template and the DSB-proximal genome sequence. As such, inhibition of the NHEJ pathway increases the chances of being repaired by MMEJ, leading to more deletions, small and large [
      • Yeh CD
      • Richardson CD
      • Corn JE.
      Advances in genome editing through control of DNA repair pathways.
      ,
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ]. Similarly, inhibition of the NHEJ pathway is a common strategy to increase HDR efficiency [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ,
      • Riesenberg S
      • Maricic T.
      Targeting repair pathways with small molecules increases precise genome editing in pluripotent stem cells.
      ,
      • Maruyama T
      • Dougan SK
      • Truttman MC
      • Bilate AM
      • Ingram JR
      • Ploegh HL.
      Increasing the efficiency of precise genome editing with CRISPR-Cas9 by inhibition of nonhomologous end joining.
      ]. Alternatively, selecting a sgRNA with lower-level NHEJ and higher-level MMEJ editing outcomes will increase relative HDR efficiency [
      • Tatiossian KJ
      • Clark RDE
      • Huang C
      • Thornton ME
      • Grubbs BH
      • Cannon PM.
      Rational selection of CRISPR-Cas9 guide RNAs for homology-directed genome editing.
      ].

      CRISPR–CAS9 INDUCES COMPLEX ON-TARGET OUTCOMES

      Large Deletions

      Recent advances in CRISPR–Cas9 technology have permitted efficient DNA modifications. Studies on the gene editing of human T cells [
      • Stadtmauer EA
      • Fraietta JA
      • Davis MM
      • et al.
      CRISPR-engineered T cells in patients with refractory cancer.
      ], hematopoietic stem cells [
      • Wu Y
      • Zeng J
      • Roscoe BP
      • et al.
      Highly efficient therapeutic gene editing of human hematopoietic stem cells.
      ], induced pluripotent stem cells [
      • Li XL
      • Li GH
      • Fu J
      • et al.
      Highly efficient genome editing via CRISPR-Cas9 in human pluripotent stem cells is achieved by transient BCL-XL overexpression.
      ], and even human embryos [
      • Ma H
      • Marti-Gutierrez N
      • Park SW
      • et al.
      Correction of a pathogenic gene mutation in human embryos.
      ] have paved the way for clinical gene therapies. In addition to CRISPR-induced small indels and template-dependent repairs at on-target sites, several unintended outcomes, such as large deletions and complex genomic rearrangements, have been reported after Cas9–sgRNA cleavages [
      • Hendel A
      • Kildebeck EJ
      • Fine EJ
      • et al.
      Quantifying genome-editing outcomes at endogenous loci with SMRT sequencing.
      ,
      • Kosicki M
      • Tomberg K
      • Bradley A.
      Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements.
      ,
      • Adikusuma F
      • Piltz S
      • Corbett MA
      • et al.
      Large deletions induced by Cas9 cleavage.
      ,
      • Song Y
      • Liu Z
      • Zhang Y
      • et al.
      Large-fragment deletions induced by Cas9 cleavage while not in the BEs system.
      ]. On-target effects, such as indels, are usually determined by sequencing a short polymerase chain reaction (PCR) product by either Sanger sequencing [
      • Brinkman EK
      • van Steensel B.
      Rapid quantitative evaluation of CRISPR genome editing by TIDE and TIDER.
      ,
      • Bloh K
      • Kanchana R
      • Bialk P
      • et al.
      Deconvolution of complex DNA repair (DECODR): establishing a novel deconvolution algorithm for comprehensive analysis of CRISPR-edited sanger sequencing data.
      ] or Illumina sequencing [
      • Park J
      • Lim K
      • Kim JS
      • Bae S.
      Cas-analyzer: an online tool for assessing genome editing results using NGS data.
      ,
      • Clement K
      • Rees H
      • Canver MC
      • et al.
      CRISPResso2 provides accurate and rapid genome editing sequence analysis.
      ]. However, the short length has eluded the detection of large deletions because large fragment deletions have depleted one or two primer-binding sites. The advances of third-generation sequencing technologies, including Oxford Nanopore and PacBio, have made it possible to reveal complex on-target mutations. These single molecular high-throughput technologies have the advantage of sequencing long nucleotides >20 kb. Coupling long-range PCR with third-generation sequencing enables quantitation of the editing alleles with long deletions [
      • Kosicki M
      • Tomberg K
      • Bradley A.
      Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements.
      ]. Deletion of many kilobases occurs after CRISPR–Cas9 gene editing in mouse and human cells. This extensive on-target genomic damage is a common outcome independent of loci and cell lines [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ,
      • Kosicki M
      • Tomberg K
      • Bradley A.
      Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements.
      ,
      • Adikusuma F
      • Piltz S
      • Corbett MA
      • et al.
      Large deletions induced by Cas9 cleavage.
      ,
      • Song Y
      • Liu Z
      • Zhang Y
      • et al.
      Large-fragment deletions induced by Cas9 cleavage while not in the BEs system.
      ]. The recent development of the GREPore-seq workflow, which combines long-range PCR with Nanopore sequencing, enables scalable and quantitative assessment of large deletions cost-effectively [Quan et al., 2021, unpublished, doi: https://doi.org/10.1101/2021.12.13.472514]. However, because of the intrinsic limitations of long-range PCR, megadeletions >10 kb have not been fully illustrated, and their occurrence may depend on the unique features of the targets. That said, deletions exceeding 1 kb are observed at considerably lower frequencies than shorter deletions (100–500 bp) [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ].

      Chromosome Rearrangement

      Chromosome rearrangement can occur when a cell simultaneously generates two or more DSBs. For example, under replication stress, fork collapse induces the formation of a DSB at the stalled fork; thus, rearrangement or translocation may sporadically arise in the presence of multiple stalled forks [
      • Howarth KD
      • Pole JCM
      • Beavis JC
      • et al.
      Large duplications at reciprocal translocation breakpoints that might be the counterpart of large deletions and could arise from stalled replication bubbles.
      ]. Double-strand cut by CRISPR–Cas9 can presumably increase chromosome rearrangement frequencies, particularly when cutting occurs at both an on-target and one or more off-targets. Simultaneously editing two or more sites will exacerbate this problem, for instance, in cases where multiple genes need to be depleted to generate universal “off-the-shelf” CAR-T cells, such as genes related to immune compatibility (B2M [
      • Gornalusse GG
      • Hirata RK
      • Funk SE
      • et al.
      HLA-E-expressing pluripotent stem cells escape allogeneic responses and lysis by NK cells.
      ], CIITA [
      • DeSandro A
      • Nagarajan UM
      • Boss JM.
      The bare lymphocyte syndrome: molecular clues to the transcriptional regulation of major histocompatibility complex class II genes.
      ], TRAC, or TRBC [
      • Torikai H
      • Reik A
      • Liu PQ
      • et al.
      A foundation for universal T-cell based immunotherapy: T cells engineered to express a CD19-specific chimeric-antigen-receptor and eliminate expression of endogenous TCR.
      ]) and immune checkpoints such as PDCD1 [
      • Ren J
      • Liu X
      • Fang C
      • Jiang S
      • June CH
      • Zhao Y.
      Multiplex genome editing to generate universal CAR T cells resistant to PD1 inhibition.
      ]. In addition, an sgRNA targeting TRBC may lead to an ∼9-kb deletion because of the presence of two proximal paralogs TRBC1 and TRBC2, with nearly identical sequences. As expected, gene engineering of several loci in human T cells leads to translocations at an ∼1% frequency [
      • Stadtmauer EA
      • Fraietta JA
      • Davis MM
      • et al.
      CRISPR-engineered T cells in patients with refractory cancer.
      ]. Although the number of cells with chromosome rearrangements drops after CAR-T cell infusion in patients, it deserves attention whether a large number of rearranged T cells in patients will engender unintended consequences.

      Loss of Heterozygosity

      Loss of heterozygosity (LOH) is a genetic alteration process in which heterozygous cells lose an allele and become homozygous or hemizygous [
      • Happle R.
      Loss of heterozygosity in human skin.
      ]. LOH is a common genetic event in cancer development and can lead to malignant growth [
      • Ryland GL
      • Doyle MA
      • Goode D
      • et al.
      Loss of heterozygosity: what is it good for?.
      ]. LOH also occurs after CRISPR gene editing [
      • Weisheit I
      • Kroeger JA
      • Malik R
      • et al.
      Detection of deleterious on-target effects after HDR-mediated CRISPR editing.
      ,
      • Alanis-Lobato G
      • Zohren J
      • McCarthy A
      • Niakan KK.
      Frequent loss of heterozygosity in CRISPR-Cas9-edited early human embryos.
      ]. Digital karyotyping and single-nucleotide polymorphism (SNP) genotyping-based tools can be used to detect megadeletions or loss of heterozygosity at the single-cell clone level. Such methods can be applied to tumor cell lines or iPSCs. In human iPSCs, large deletions and loss of heterozygosity were reported to occur in up to 40% of edited clones in one study [
      • Weisheit I
      • Kroeger JA
      • Malik R
      • et al.
      Detection of deleterious on-target effects after HDR-mediated CRISPR editing.
      ]. However, another study indicated few large deletions and undetectable loss of heterozygosity in human iPSCs after CRISPR cleavage [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ]. The conflicting results on large deletion frequencies after CRISPR editing may be attributed to the different CRISPR delivery approaches and/or detection methods.

      Chromothripsis and Chromosome Loss

      Chromothripsis is a process defined as chromosome shattering with chromosome structural abnormalities, which is common in cancer and can generate fusion oncogenes or mutation of tumor suppressors [
      • Forment JV
      • Kaidi A
      • Jackson SP.
      Chromothripsis and cancer: causes and consequences of chromosome shattering.
      ,
      • Umbreit NT
      • Zhang CZ
      • Lynch LD
      • et al.
      Mechanisms generating cancer genome complexity from a single cell division error.
      ]. A recent study reported that chromothripsis is an unappreciated on-target outcome after CRISPR–Cas9-induced DSBs [
      • Leibowitz ML
      • Papathanasiou S
      • Doerfler PA
      • et al.
      Chromothripsis as an on-target consequence of CRISPR–Cas9 genome editing.
      ]. In dividing cells, this process includes the formation of micronuclei and chromosome bridges, thus leading to large chromosomal segment loss and chromothripsis, which is also a potential mechanism of LOH mentioned above. However, whether large deletions accompany LOH or chromothripsis has not been revealed. The mechanisms underlying editing-triggered LOH or chromothripsis are pressing questions that need to be addressed in future investigations.
      In addition to micronucleus formation and chromothripsis, whole-chromosome loss was observed after CRISPR–Cas9 genome editing of early embryos, which can be propagated over several divisions of embryonic development [
      • Papathanasiou S
      • Markoulaki S
      • Blaine LJ
      • et al.
      Whole chromosome loss and genomic instability in mouse embryos after CRISPR–Cas9 genome editing.
      ,
      • Zuccaro MV
      • Xu J
      • Mitchell C
      • et al.
      Allele-specific chromosome removal after Cas9 cleavage in human embryos.
      ]. Although chromosome loss is a serious genetic alteration after genomic editing, it is reported only in human and mouse embryos thus far, likely associated with the giant size of cells during early development. It is still unknown whether it can occur in primary and stem cells of clinical interest.

      Challenges Facing Clinical Genomic Editing

      As CRISPR–Cas9 gene editing can result in complex on-target outcomes and off-target effects (Figure 1), this field calls for a standardized methodology to assess gene editing outcomes comprehensively [
      • Lazzarotto CR
      • Malinin NL
      • Li Y
      • et al.
      CHANGE-seq reveals genetic and epigenetic effects on CRISPR–Cas9 genome-wide activity.
      ]. Thus far, off-target prediction and detection methods have been well developed [
      • Tsai SQ
      • Zheng Z
      • Nguyen NT
      • et al.
      GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases.
      ,
      • Lazzarotto CR
      • Malinin NL
      • Li Y
      • et al.
      CHANGE-seq reveals genetic and epigenetic effects on CRISPR–Cas9 genome-wide activity.
      ]. However, few protocols are in place to assess chromosome rearrangement, LOH, and chromothripsis precisely. Recently, in a preprint, Nahmad et al. [unpublished, 2021, doi: https://doi.org/10.1101/2021.08.20.457092] reported that genomic editing of clinically relevant loci in human primary T cells leads to complex chromosome truncations and aneuploidy by single-cell RNA sequencing (scRNA-seq) These data suggest that we may harness the single-cell sequencing technology to quality control the editing products. However, it is still especially challenging during the transition from bench to bedside, as clinical genomic editing and cell therapy need to engineer cells on the order of magnitude of millions to billions. As such, the presence of 0.1% undesired mutations in the infusion products can translate to 105 to 107 mutated cells in the body, likely eliciting pathogenic consequences.
      Figure 1
      Figure 1CRISPR–Cas9 induced complex genomic editing outcomes. (A) On-target double-strand breaks (DSB) repair through NHEJ or MMEJ will lead to small indels. (B) Delayed or failed repair by NHEJ or MMEJ may result in large deletions. (C–F) At the chromosome level, more complex variants, such as loss-of-heterozygosity (LOH) (C), chromosome rearrangement (D), chromothripsis (E), or even whole chromosome loss (F) will occur during CRISPR–Cas9-mediated genomic editing. (G) Nonspecific targeting of Cas9-sgRNA will lead to off-target cleavage, increasing the chance of chromosome rearrangement.

      SAFETY EVALUATION OF UNWANTED ON-TARGET MUTATIONS

      Pathogenic Consequences

      The loci for gene editing are usually transcriptionally active. Translocation of these loci may cause fusion oncogene formation and high-level expression. In the context of hematopoietic stem and progenitor cell (HSPC) transplantation, billions of cells are edited. Oncogene activation or fusion will endow cells with a proliferative advantage, especially HSPCs with a longer life span than mature somatic cells. Such an event will increase the risk of clonal hematopoiesis and leukemogenesis. In early gene therapy trials, viral vector transduction-induced insertions activated LMO2 and caused cancer in patients [
      • Hacein-Bey-Abina S
      • von Kalle C
      • Schmidt M
      • et al.
      A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency.
      ]. In current clinical gene therapies with CRISPR–Cas9-edited cells, few studies have found the pathogenic consequences of edited cells caused by the limited follow-up period. In addition, large deletions spanning kilobases or more tend to impair the function of tumor suppressors. However, whether observed genomic damage will have pathogenic effects in the clinic remains unanswered.

      Regulation of Large Deletions

      The failed or delayed repair of DSBs induced by CRISPR–Cas9 or aberrant rejoining may lead to large genomic deletions or rearrangements [
      • Zhu C
      • Mills KD
      • Ferguson DO
      • et al.
      Unrepaired DNA breaks in p53-deficient cells lead to oncogenic gene amplification subsequent to translocations.
      ]. When the NHEJ pathway is impaired, the occurrence of large deletions increases [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ]. Conversely, the frequency of large deletions decreases in cells with a deficiency in resection genes or on inhibition of MMEJ [Kosicki et al., 2021, unpublished, doi: https://doi.org/10.1101/2020.08.05.216739]. With the provision of an HDR donor template, timely and efficient HDR repair considerably attenuates large deletions [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ]. The putative mechanism might be the differential kinetics of NHEJ, HDR, or MMEJ repair machinery. As NHEJ and HDR occur more rapidly than MMEJ [
      • Fu YW
      • Dai XY
      • Wang WT
      • et al.
      Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
      ], timely repair of DSBs by NHEJ or HDR attenuates large deletions. On the other hand, the consecutive failure to repair DSBs by NHEJ, HDR, and MMEJ will eventually lead to significant mutations, including large deletions.
      Several NHEJ inhibitors, such as M3814 [
      • Haapaniemi E
      • Botla S
      • Persson J
      • Schmierer B
      • Taipale J.
      CRISPR–Cas9 genome editing induces a p53-mediated DNA damage response.
      ] or their combination with several inhibitors [
      • Maruyama T
      • Dougan SK
      • Truttman MC
      • Bilate AM
      • Ingram JR
      • Ploegh HL.
      Increasing the efficiency of precise genome editing with CRISPR-Cas9 by inhibition of nonhomologous end joining.
      ], have been used to achieve highly efficient knockin by HDR. In addition, at least one company has attempted to use NHEJ inhibitors to improve precise editing outcomes in clinical gene therapy [
      Vertex ramps up CRISPR repair.
      ]. However, because inhibition of NHEJ may increase large deletion mutations [
      • Wen W
      • Quan ZJ
      • Li SA
      • et al.
      Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
      ], the clinical use of NHEJ inhibitors to increase HDR requires a careful safety assessment. From our perspective, to achieve both a higher HDR knockin efficiency and a lower large deletion rate, we prefer to choose an sgRNA targeting site with predominant MMEJ repair tendencies and provide enough HDR donors in a timely manner.
      Other strategies to enhance HDR editing outcomes include TP53 interference and HDAC inhibition. In cells of clinical significance, such as hematopoietic stem cells and T cells, delayed DSB repair will lead to TP53 activation and cell cycle arrest [
      • Haapaniemi E
      • Botla S
      • Persson J
      • Schmierer B
      • Taipale J.
      CRISPR–Cas9 genome editing induces a p53-mediated DNA damage response.
      ,
      • Lee BC
      • Lozano RJ
      • Dunbar CE.
      Understanding and overcoming adverse consequences of genome editing on hematopoietic stem and progenitor cells.
      ]. In iPSCs, inhibition of TP53 can increase the efficiency of HDR-mediated precise knockin [
      • Ihry RJ
      • Worringer KA
      • Salick MR
      • et al.
      p53 inhibits CRISPR–Cas9 engineering in human pluripotent stem cells.
      ]. However, disruption of TP53 may cause genomic instability and elevate deleterious mutation events. HDAC inhibition builds up histone acetylation, leading to chromatin decondensation and escalating accessibility to editing components. These changes result in higher NHEJ- or MMEJ-mediated indels and HDR-mediated knockin at both open and closed loci, with the most marked effects observed with HDR at closed loci [
      • Zhang J-P
      • Yang ZX
      • Zhang F
      • et al.
      HDAC inhibitors improve CRISPR-mediated HDR editing efficiency in iPSCs.
      ]. However, the impact of HDAC inhibitors on the generation of unwanted on-target mutations has not been reported.

      CRISPR–Cas9-Mediated Gene Editing Without Double-Strand Breaks

      Most CRISPR–Cas9 editing studies currently rely on the generation of DNA DSBs, which may induce chromatin instability and unintended consequences (Figure 1). To considerably reduce the harmful effects of gene editing, several approaches, such as base editing (BE) and primer editing (PE), modify genes by cutting only one strand of DNA [
      • Anzalone AV
      • Koblan LW
      • Liu DR.
      Genome editing with CRISPR-Cas nucleases, base editors, transposases and prime editors.
      ]. Base editors use deaminase to achieve alteration of a single nucleotide, such as A to G or C to T, using a Cas9 nickase without inducing DSBs [
      • Komor AC
      • Kim YB
      • Packer MS
      • Zuris JA
      • Liu DR.
      Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage.
      ]. Thus, the BE tool has promise for applications in gene therapy of single-nucleotide-mutated diseases. PE technology fuses a Cas9 nickase with a reverse transcriptase using a short template provided by pegRNA to guide the insertion or deletion of a short fragment into the target site [
      • Anzalone AV
      • Randolph PB
      • Davis JR
      • et al.
      Search-and-replace genome editing without double-strand breaks or donor DNA.
      ]. PE enables precise editing of a short DNA fragment (usually multiple bases) [
      • Chen PJ
      • Hussmann JA
      • Yan J
      • et al.
      Enhanced prime editing systems by manipulating cellular determinants of editing outcomes.
      ]. However, both BE and PE, in their current version, cannot knock in a long DNA fragment, even though the optimized PE system has succeeded in deleting a large piece [
      • Jiang T
      • Zhang XO
      • Weng Z
      • Xue W
      Deletion and replacement of long genomic sequences using prime editing.
      ,
      • Choi J
      • Chen W
      • Suiter CC
      • et al.
      Precise genomic deletions using paired prime editing.
      ]. It is worth mentioning that the optimized PE system will also introduce DSBs, albeit at low frequency. Even with potential limitations, CRISPR-based BE or PE technologies are finding applications in clinical gene editing. However, because of a small number of DSBs in BE- or PE-enabled editing, assessment of adverse events is still necessary.

      Perspectives on Safe Clinical Gene Therapy

      With the development of a staggering variety of gene editing tools and the impressive progress of clinical cell therapy trials, gene therapy has become a powerful tool to treat genetic disorders or endow cells with the capacity to battle diseases. Gene therapy provides a cure for thousands of genetic disorders [
      • Landrum MJ
      • Lee JM
      • Riley GR
      • et al.
      ClinVar: public archive of relationships among sequence variation and human phenotype.
      ]. Precise genomic editing offers a way to repair disrupted genomic sites and treat various diseases. However, one should keep in mind off-target editing and other unintended consequences to ensure clinical safety. The off-target activity can be minimized or reduced using Hi-Fi Cas9 enzymes [
      • Kleinstiver BP
      • Pattanayak V
      • Prew MS
      • et al.
      High-fidelity CRISPR–Cas9 nucleases with no detectable genome-wide off-target effects.
      ,
      • Slaymaker IM
      • Gao L
      • Zetsche B
      • Scott DA
      • Yan WX
      • Zhang F.
      Rationally engineered Cas9 nucleases with improved specificity.
      ] and optimized delivery approaches [
      • Tsai SQ
      • Joung JK.
      Defining and improving the genome-wide specificities of CRISPR–Cas9 nucleases.
      ].
      Investigators have focused largely on improving on-target editing efficiency to obtain an ideal result for clinical gene therapy. Nevertheless, few studies have paid sufficient attention to detecting comprehensive on-target editing consequences, largely because of technical challenges. With the surge in gene therapy clinical trials, preclinical safety evaluation criteria and long-term follow-up guidelines must be instituted. In addition to the proper quality controls and detection methods that have been implemented, we also need to develop greater editing tools and optimize experimental procedures to attain more precise gene editing (Table 1). Furthermore, a continuous investigation into the mechanisms underlying DNA repair, such as the kinetics and relationships of DNA repair pathways, is imperative.
      Table 1Potential solutions to controlling complex unwanted outcomes
      On-target outcomePotential solution
      Large deletionsBase editing, prime editing, HDR editing, NHEJ-dsODN insertion, etc.
      Loss of heterozygosity
      Chromosome rearrangement
      Chromothripsis
      Chromosome loss

      Conflict of Interest Disclosure

      The authors declare that they have no competing interests.

      Acknowledgments

      This work was supported by grants from the CAMS Innovation Fund for Medical Sciences (CIFMS) (2021-I2M-1-041); Tianjin Synthetic Biotechnology Innovation Capacity Improvement Project (TSBICIP-KJGG-017); and Nonprofit Central Research Institute Fund of the Chinese Academy of Medical Sciences (2020-PT310-011).

      References

        • Lander ES.
        The heroes of CRISPR.
        Cell. 2016; 164: 18-28
        • Deltcheva E
        • Chylinski K
        • Sharma CM
        • et al.
        CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III.
        Nature. 2011; 471: 602-607
        • Gasiunas G
        • Barrangou R
        • Horvath P
        • Siksnys V.
        Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria.
        Proc Natl Acad Sci USA. 2012; 109: E2579-E2586
        • Jinek M
        • Chylinski K
        • Fonfara I
        • Hauer M
        • Doudna JA
        • Charpentier E.
        A programmable dual-RNA–guided DNA endonuclease in adaptive bacterial immunity.
        Science. 2012; 337: 816-821
        • Nishimasu H
        • Ran FA
        • Hsu PD
        • et al.
        Crystal structure of Cas9 in complex with guide RNA and target DNA.
        Cell. 2014; 156: 935-949
        • Cong L
        • Ran FA
        • Cox D
        • et al.
        Multiplex genome engineering using CRISPR/Cas systems.
        Science. 2013; 339: 819-823
        • Komor AC
        • Kim YB
        • Packer MS
        • Zuris JA
        • Liu DR.
        Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage.
        Nature. 2016; 533: 420-424
        • Gaudelli NM
        • Komor AC
        • Rees HA
        • et al.
        Programmable base editing of A•T to G•C in genomic DNA without DNA cleavage.
        Nature. 2017; 551: 464-471
        • Rees HA
        • Liu DR.
        Base editing: precision chemistry on the genome and transcriptome of living cells.
        Nat Rev Genet. 2018; 19: 770-788
        • Anzalone AV
        • Randolph PB
        • Davis JR
        • et al.
        Search-and-replace genome editing without double-strand breaks or donor DNA.
        Nature. 2019; 576: 149-157
        • Shalem O
        • Sanjana NE
        • Zhang F.
        High-throughput functional genomics using CRISPR–Cas9.
        Nat Rev Genet. 2015; 16: 299-311
        • Thakore PI
        • Black JB
        • Hilton IB
        • Gersbach CA.
        Editing the epigenome: technologies for programmable transcription and epigenetic modulation.
        Nat Methods. 2016; 13: 127-137
        • Pickar-Oliver A
        • Gersbach CA.
        The next generation of CRISPR–Cas technologies and applications.
        Nat Rev Mol Cell Biol. 2019; 20: 490-507
        • Shou J
        • Li J
        • Liu Y
        • Wu Q.
        Precise and predictable CRISPR chromosomal rearrangements reveal principles of Cas9-mediated nucleotide insertion.
        Mol Cell. 2018; 71 (e4): 498-509
        • Hussmann JA
        • Ling J
        • Ravisankar P
        • et al.
        Mapping the genetic landscape of DNA double-strand break repair.
        Cell. 2021; 184 (e25): 5653-5669
        • Nambiar TS
        • Baudrier L
        • Billon P
        • Ciccia A.
        CRISPR-based genome editing through the lens of DNA repair.
        Mol Cell. 2022; 82: 348-388
        • Scully R
        • Panday A
        • Elango R
        • Willis NA.
        DNA double-strand break repair-pathway choice in somatic mammalian cells.
        Nat Rev Mol Cell Biol. 2019; 20: 698-714
        • Brinkman EK
        • Chen T
        • de Haas M
        • Holland HA
        • Akhtar W
        • van Steensel B.
        Kinetics and fidelity of the repair of Cas9-induced double-strand DNA breaks.
        Mol Cell. 2018; 70 (e6): 801-813
        • Wang T
        • Wei JJ
        • Sabatini DM
        • Lander ES.
        Genetic screens in human cells using the CRISPR-Cas9 system.
        Science. 2014; 343: 80-84
        • Shalem O
        • Sanjana NE
        • Hartenian E
        • et al.
        Genome-scale CRISPR-Cas9 knockout screening in human cells.
        Science. 2014; 343: 84-87
        • Brogna S
        • Wen J.
        Nonsense-mediated mRNA decay (NMD) mechanisms.
        Nat Struct Mol Biol. 2009; 16: 107-113
        • Popp MW
        • Maquat LE.
        Leveraging rules of nonsense-mediated mRNA decay for genome engineering and personalized medicine.
        Cell. 2016; 165: 1319-1322
        • van Overbeek M
        • Capurso D
        • Carter MM
        • et al.
        DNA repair profiling reveals nonrandom outcomes at Cas9-mediated breaks.
        Mol Cell. 2016; 63: 633-646
        • Allen F
        • Crepaldi L
        • Alsinet C
        • et al.
        Predicting the mutations generated by repair of Cas9-induced double-strand breaks.
        Nat Biotechnol. 2019; 37: 64-72
        • Fu YW
        • Dai XY
        • Wang WT
        • et al.
        Dynamics and competition of CRISPR-Cas9 ribonucleoproteins and AAV donor-mediated NHEJ, MMEJ and HDR editing.
        Nucleic Acids Res. 2021; 49: 969-985
        • Zhang JP
        • Cheng XX
        • Zhao M
        • et al.
        Curing hemophilia A by NHEJ-mediated ectopic F8 insertion in the mouse.
        Genome Biol. 2019; 20: 276
        • Auer TO
        • Duroure K
        • De Cian A
        • Concordet JP
        • Del Bene F.
        Highly efficient CRISPR/Cas9-mediated knock-in in zebrafish by homology-independent DNA repair.
        Genome Res. 2014; 24: 142-153
        • Suzuki K
        • Tsunekawa Y
        • Hernandez-Benitez R
        • et al.
        In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration.
        Nature. 2016; 540: 144-149
        • Tsai SQ
        • Zheng Z
        • Nguyen NT
        • et al.
        GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases.
        Nat Biotechnol. 2015; 33: 187-197
        • Sfeir A
        • Symington LS.
        Microhomology-mediated end joining: a back-up survival mechanism or dedicated pathway?.
        Trends Biochem Sci. 2015; 40: 701-714
        • Iyer S
        • Suresh S
        • Guo D
        • et al.
        Precise therapeutic gene correction by a simple nuclease-induced double-stranded break.
        Nature. 2019; 568: 561-565
        • Wang L
        • Li L
        • Ma Y
        • et al.
        Reactivation of γ-globin expression through Cas9 or base editor to treat β-hemoglobinopathies.
        Cell Res. 2020; 30: 276-278
        • Porteus MH.
        Towards a new era in medicine: therapeutic genome editing.
        Genome Biol. 2015; 16: 1-12
        • Dever DP
        • Porteus MH.
        The changing landscape of gene editing in hematopoietic stem cells: a step towards Cas9 clinical translation.
        Curr Opinion Hematol. 2017; 24: 481
        • Chu VT
        • Weber T
        • Wefers B
        • et al.
        Increasing the efficiency of homology-directed repair for CRISPR–Cas9-induced precise gene editing in mammalian cells.
        Nat Biotechnol. 2015; 33: 543-548
        • Takata M
        • Sasaki MS
        • Sonoda E
        • et al.
        Homologous recombination and non-homologous end-joining pathways of DNA double-strand break repair have overlapping roles in the maintenance of chromosomal integrity in vertebrate cells.
        EMBO J. 1998; 17: 5497-5508
        • Li XL
        • Li GH
        • Fu J
        • et al.
        Highly efficient genome editing via CRISPR-Cas9 in human pluripotent stem cells is achieved by transient BCL-XL overexpression.
        Nucl Acids Res. 2018; 46: 10195-10215
        • Zhang JP
        • Li XL
        • Li GH
        • et al.
        Efficient precise knockin with a double cut HDR donor after CRISPR/Cas9-mediated double-stranded DNA cleavage.
        Genome Biol. 2017; 18: 35
        • Wen W
        • Cheng X
        • Fu Y
        • et al.
        High-level precise knockin of iPSCs by simultaneous reprogramming and genome editing of human peripheral blood mononuclear cells.
        Stem Cell Rep. 2018; 10: 1821-1834
        • Kuo CY
        • Long JD
        • Campo-Fernandez B
        • et al.
        Site-specific gene editing of human hematopoietic stem cells for X-linked hyper-IgM syndrome.
        Cell Rep. 2018; 23: 2606-2616
        • Vakulskas CA
        • Dever DP
        • Rettig GR
        • et al.
        A high-fidelity Cas9 mutant delivered as a ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells.
        Nat Med. 2018; 24: 1216-1224
        • Pavel-Dinu M
        • Wiebking V
        • Dejene BT
        • et al.
        Gene correction for SCID-X1 in long-term hematopoietic stem cells.
        Nat Commun. 2019; 10: 1634
        • Ferrari S
        • Jacob A
        • Beretta S
        • et al.
        Efficient gene editing of human long-term hematopoietic stem cells validated by clonal tracking.
        Nat Biotechnol. 2020; 38: 1298-1308
        • Quadros RM
        • Miura H
        • Harms DW
        • et al.
        Easi-CRISPR: a robust method for one-step generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR ribonucleoproteins.
        Genome Biol. 2017; 18: 92
        • Tatiossian KJ
        • Clark RDE
        • Huang C
        • Thornton ME
        • Grubbs BH
        • Cannon PM.
        Rational selection of CRISPR-Cas9 guide RNAs for homology-directed genome editing.
        Mol Ther. 2021; 29: 1057-1069
        • Roidos P
        • Sungalee S
        • Benfatto S
        • et al.
        A scalable CRISPR/Cas9-based fluorescent reporter assay to study DNA double-strand break repair choice.
        Nat Commun. 2020; 11: 4077
        • Truong LN
        • Li Y
        • Shi LZ
        • et al.
        Microhomology-mediated end joining and homologous recombination share the initial end resection step to repair DNA double-strand breaks in mammalian cells.
        Proc Natl Acad Sci USA. 2013; 110: 7720-7725
        • Yeh CD
        • Richardson CD
        • Corn JE.
        Advances in genome editing through control of DNA repair pathways.
        Nat Cell Biol. 2019; 21: 1468-1478
        • Wen W
        • Quan ZJ
        • Li SA
        • et al.
        Effective control of large deletions after double-strand breaks by homology-directed repair and dsODN insertion.
        Genome Biol. 2021; 22: 236
        • Riesenberg S
        • Maricic T.
        Targeting repair pathways with small molecules increases precise genome editing in pluripotent stem cells.
        Nat Commun. 2018; 9: 2164
        • Maruyama T
        • Dougan SK
        • Truttman MC
        • Bilate AM
        • Ingram JR
        • Ploegh HL.
        Increasing the efficiency of precise genome editing with CRISPR-Cas9 by inhibition of nonhomologous end joining.
        Nat Biotechnol. 2015; 33: 538-542
        • Stadtmauer EA
        • Fraietta JA
        • Davis MM
        • et al.
        CRISPR-engineered T cells in patients with refractory cancer.
        Science. 2020; 367: eaba7365
        • Wu Y
        • Zeng J
        • Roscoe BP
        • et al.
        Highly efficient therapeutic gene editing of human hematopoietic stem cells.
        Nat Med. 2011; 25: 776-783
        • Ma H
        • Marti-Gutierrez N
        • Park SW
        • et al.
        Correction of a pathogenic gene mutation in human embryos.
        Nature. 2017; 548: 413-419
        • Hendel A
        • Kildebeck EJ
        • Fine EJ
        • et al.
        Quantifying genome-editing outcomes at endogenous loci with SMRT sequencing.
        Cell Rep. 2014; 7: 293-305
        • Kosicki M
        • Tomberg K
        • Bradley A.
        Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements.
        Nat Biotechnol. 2018; 36: 765-771
        • Adikusuma F
        • Piltz S
        • Corbett MA
        • et al.
        Large deletions induced by Cas9 cleavage.
        Nature. 2018; 560: E8-E9
        • Song Y
        • Liu Z
        • Zhang Y
        • et al.
        Large-fragment deletions induced by Cas9 cleavage while not in the BEs system.
        Mol Ther Nucl Acids. 2020; 21: 523-526
        • Brinkman EK
        • van Steensel B.
        Rapid quantitative evaluation of CRISPR genome editing by TIDE and TIDER.
        in: Luo Y CRISPR Gene Editing. Humana Press, New York1961: 29-44
        • Bloh K
        • Kanchana R
        • Bialk P
        • et al.
        Deconvolution of complex DNA repair (DECODR): establishing a novel deconvolution algorithm for comprehensive analysis of CRISPR-edited sanger sequencing data.
        CRISPR J. 2021; 4: 120-131
        • Park J
        • Lim K
        • Kim JS
        • Bae S.
        Cas-analyzer: an online tool for assessing genome editing results using NGS data.
        Bioinformatics. 2017; 33: 286-288
        • Clement K
        • Rees H
        • Canver MC
        • et al.
        CRISPResso2 provides accurate and rapid genome editing sequence analysis.
        Nat Biotechnol. 2019; 37: 224-226
        • Howarth KD
        • Pole JCM
        • Beavis JC
        • et al.
        Large duplications at reciprocal translocation breakpoints that might be the counterpart of large deletions and could arise from stalled replication bubbles.
        Genome Res. 2011; 21: 525-534
        • Gornalusse GG
        • Hirata RK
        • Funk SE
        • et al.
        HLA-E-expressing pluripotent stem cells escape allogeneic responses and lysis by NK cells.
        Nat Biotechnol. 2017; 35: 765-772
        • DeSandro A
        • Nagarajan UM
        • Boss JM.
        The bare lymphocyte syndrome: molecular clues to the transcriptional regulation of major histocompatibility complex class II genes.
        Am J Hum Genet. 1999; 65: 279-286
        • Torikai H
        • Reik A
        • Liu PQ
        • et al.
        A foundation for universal T-cell based immunotherapy: T cells engineered to express a CD19-specific chimeric-antigen-receptor and eliminate expression of endogenous TCR.
        Blood. 2012; 119: 5697-5705
        • Ren J
        • Liu X
        • Fang C
        • Jiang S
        • June CH
        • Zhao Y.
        Multiplex genome editing to generate universal CAR T cells resistant to PD1 inhibition.
        Clin Cancer Res. 2017; 23: 2255-2266
        • Happle R.
        Loss of heterozygosity in human skin.
        J Am Acad Dermatol. 1999; 41: 143-161
        • Ryland GL
        • Doyle MA
        • Goode D
        • et al.
        Loss of heterozygosity: what is it good for?.
        BMC Med Genomics. 2015; 8: 1-12
        • Weisheit I
        • Kroeger JA
        • Malik R
        • et al.
        Detection of deleterious on-target effects after HDR-mediated CRISPR editing.
        Cell Rep. 2020; 31107689
        • Alanis-Lobato G
        • Zohren J
        • McCarthy A
        • Niakan KK.
        Frequent loss of heterozygosity in CRISPR-Cas9-edited early human embryos.
        Proc Natl Acad Sci USA. 2021; 118e2004832117
        • Forment JV
        • Kaidi A
        • Jackson SP.
        Chromothripsis and cancer: causes and consequences of chromosome shattering.
        Nat Rev Cancer. 2012; 12: 663-670
        • Umbreit NT
        • Zhang CZ
        • Lynch LD
        • et al.
        Mechanisms generating cancer genome complexity from a single cell division error.
        Science. 2020; 368: eaba0712
        • Leibowitz ML
        • Papathanasiou S
        • Doerfler PA
        • et al.
        Chromothripsis as an on-target consequence of CRISPR–Cas9 genome editing.
        Nat Genet. 2021; 53: 895-905
        • Papathanasiou S
        • Markoulaki S
        • Blaine LJ
        • et al.
        Whole chromosome loss and genomic instability in mouse embryos after CRISPR–Cas9 genome editing.
        Nat Commun. 2021; 12: 1-7
        • Zuccaro MV
        • Xu J
        • Mitchell C
        • et al.
        Allele-specific chromosome removal after Cas9 cleavage in human embryos.
        Cell. 2020; 183 (e15): 1650-1664
        • Lazzarotto CR
        • Malinin NL
        • Li Y
        • et al.
        CHANGE-seq reveals genetic and epigenetic effects on CRISPR–Cas9 genome-wide activity.
        Nat Biotechnol. 2020; 38: 1317-1327
        • Hacein-Bey-Abina S
        • von Kalle C
        • Schmidt M
        • et al.
        A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency.
        N Engl J Med. 2003; 348: 255-256
        • Zhu C
        • Mills KD
        • Ferguson DO
        • et al.
        Unrepaired DNA breaks in p53-deficient cells lead to oncogenic gene amplification subsequent to translocations.
        Cell. 2002; 109: 811-821
        • Haapaniemi E
        • Botla S
        • Persson J
        • Schmierer B
        • Taipale J.
        CRISPR–Cas9 genome editing induces a p53-mediated DNA damage response.
        Nat Med. 2018; 24: 927-930
      1. Vertex ramps up CRISPR repair.
        Nat Biotechnol. 2019; 37: 205
        • Lee BC
        • Lozano RJ
        • Dunbar CE.
        Understanding and overcoming adverse consequences of genome editing on hematopoietic stem and progenitor cells.
        Mol Ther. 2021; 29: 3205-3218
        • Ihry RJ
        • Worringer KA
        • Salick MR
        • et al.
        p53 inhibits CRISPR–Cas9 engineering in human pluripotent stem cells.
        Nat Med. 2018; 24: 939-946
        • Zhang J-P
        • Yang ZX
        • Zhang F
        • et al.
        HDAC inhibitors improve CRISPR-mediated HDR editing efficiency in iPSCs.
        Sci China Life Sci. 2021; 64: 1449-1462
        • Anzalone AV
        • Koblan LW
        • Liu DR.
        Genome editing with CRISPR-Cas nucleases, base editors, transposases and prime editors.
        Nat Biotechnol. 2020; 38: 824-844
        • Chen PJ
        • Hussmann JA
        • Yan J
        • et al.
        Enhanced prime editing systems by manipulating cellular determinants of editing outcomes.
        Cell. 2021; 184 (e29): 5635-5652
        • Jiang T
        • Zhang XO
        • Weng Z
        • Xue W
        Deletion and replacement of long genomic sequences using prime editing.
        Nat Biotechnol. 2022; 40: 227-234
        • Choi J
        • Chen W
        • Suiter CC
        • et al.
        Precise genomic deletions using paired prime editing.
        Nat Biotechnol. 2022; 40: 218-226
        • Landrum MJ
        • Lee JM
        • Riley GR
        • et al.
        ClinVar: public archive of relationships among sequence variation and human phenotype.
        Nucl Acids Res. 2014; 42: D980-D985
        • Kleinstiver BP
        • Pattanayak V
        • Prew MS
        • et al.
        High-fidelity CRISPR–Cas9 nucleases with no detectable genome-wide off-target effects.
        Nature. 2016; 529: 490-495
        • Slaymaker IM
        • Gao L
        • Zetsche B
        • Scott DA
        • Yan WX
        • Zhang F.
        Rationally engineered Cas9 nucleases with improved specificity.
        Science. 2016; 351: 84-88
        • Tsai SQ
        • Joung JK.
        Defining and improving the genome-wide specificities of CRISPR–Cas9 nucleases.
        Nat Rev Genet. 2016; 17: 300-312