Regulations and safety assessment of genome editing technologies for human gene therapy

Abstract

Genome editing technologies, including clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated protein 9 (Cas9) system, are expected to become a state-of-the-art strategy for human gene therapy. However, there are several new safety issues to be addressed before clinical use. This review article summarizes the current regulatory status of the safety assessment of genome editing for human gene therapy. In addition, with a focus on unintended genome editing by CRISPR/Cas9 system, this review outlines the methods used to detect off-target sites, and introduces a proposed technical guidance on the safety assessment of ex vivo genome-edited cell products that have been developed by the Regulatory Science (RS) research project of the Japan Agency for Medical Research and Development (AMED) for genome editing.

Introduction

Genome editing technologies have great potential as tools to facilitate gene therapy for hereditary diseases, by the destruction or repair of the responsible genes. It can also be used to develop therapies that are not amenable to conventional gene therapy, for instance, the universalization of allogeneic therapeutic cells such as universal chimeric antigen receptor (CAR) T cells. The genome editing technologies currently in clinical trials include zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), and CRISPR/Cas9 system. Each of these genome editing tools specifically binds to target DNA sequences and introduces double-strand break (DSB) at the specific target site, followed by genome editing using the DNA-repair mechanism of cells. However, this type of genome editing mechanism has specific safety issues that differ from conventional gene therapy, with one of the most important issues being off-target genome editing.

This review summarizes the safety issues specific to genome editing and the current regulatory landscape of genome editing for human gene therapy. In addition, with a focus on the assessment of unintended genome editing by CRISPR/Cas9 system, this article describes the proposed guidance on the safety assessment of genome-edited cells, developed by the AMED RS research project for genome editing.

Safety Issues Specific to Genome Editing Technologies for Human Gene Therapy

Risk of off-target mutations

The genome editing technology edits specific genes by sequence-specific cleavage, however, there is a risk of editing similar sequences, also called off-target effects. The CRISPR/Cas9 system consists of a Cas9 nuclease complexed with a single guide RNA (sgRNA), the latter containing a complementary sequence of the target DNA sequence. The sgRNA binds to the target sequence by complementary binding. However, sgRNA may also bind to similar unintended DNA sequences with mismatches or insertions and deletions (INDELs), resulting in off-target editing.

Of particular concern is the tumorigenesis of cells, which could be a consequence of this off-target editing. The off-target effects may result in direct oncogene activation or inactivation of cancer suppressor genes. Moreover, gene modification by genome editing has a permanent irreversible effect. Therefore, a risk assessment of off-target mutations is essential before clinical use.

To reduce the risk of off-target mutations, it is important to design sgRNAs with few candidate off-target sites and highly specific sequences. When designing highly specific sgRNAs, in silico analysis using sequence search software such as GGGenome [1] or guide RNA design software such as CRISPRdirect [2] is useful.

Risk of unintended mutation at on-target cleavage site

DSBs induced by the CRISPR/Cas9 system are usually repaired by non-homologous end joining (NHEJ), which can be accompanied by INDELs ranging from a few to dozens of bases, or by homology-directed repair (HDR) when homologous sequences are present in the cell [3]. Additionally, however, genome editing can also result in unintended insertions of large DNA sequences of tens to thousands of nucleotides [4], or large deletions and even major changes in the genome such as inversions and chromosomal translocations [5]. These are undesired mutations at the on-target site. Unintended large insertion sequences have been observed to be DNA fragments derived from vectors introduced into the cells for genome editing, as well as DNA fragments derived from endogenous retrotransposons and mRNAs [4]. It has also been reported that once DSBs are introduced, undesired mutation may occur not only at the on-target sites, but also at the off-target sites [6,7,8]. The risk of tumorigenesis due to these chromosomal aberrations should also be taken into consideration.

Risk of P53 mutation

CRISPR/Cas9-induced DSB activates p53-mediated DNA damage response and cell cycle arrest [9], and could even be toxic to pluripotent stem cells [10]. Inhibition of the tumor suppressor p53 prevents the damage response and increases the rate of homologous recombination from a donor template [9]. Since these results suggest that precisely corrected cells using CRISPR/Cas9 may contain mutated p53, it is necessary to evaluate the p53 function of genome-edited cells.

Regulation of Gene Therapy and Genome Editing in the United States, Europe and Japan

In the United States, Europe, and Japan, several guidelines for gene therapy have been revised to take genome editing into account (Table 1).

Table 1. Guidelines for gene therapy and genome editing therapy in US, EU and Japan

	Guideline	Description of genome editing
EU EMA	Guideline on the quality, non-clinical and clinical aspects of gene therapy medicinal products (Mar. 2018) [11].	Genome editing tools are included in gene therapy products.Does not include description specific to genome editing.
	Guideline on quality, non-clinical and clinical aspects of medicinal products containing genetically modified cells (Draft, Jul. 2018) [12].	Detailed description for quality and safety of ex vivo genome-edited cells.
US FDA	Chemistry, Manufacturing, and Control (CMC) Information for Human Gene Therapy Investigational New Drug Applications (INDs), Guidance for Industry (Jan. 2020) [13].	Genome editing therapy is defined as gene therapy.Does not include description specific to safety of genome editing.
	Long Term Follow-Up After Administration of Human Gene Therapy Products, Guidance for Industry (Jan. 2020) [14].	Genome editing requires long term follow-up because of the permanent alteration of the genome.
Japan MHLW PMDA	Guidelines for Gene Therapy Clinical Research (Feb 2019) [15].	Genome editing therapy is defined as gene therapy.Does not include description specific to genome editing.
	Guideline on Ensuring the Quality and Safety of Gene Therapy Products (Jul. 2019) [16].	Guideline for conventional in vivo and ex vivo gene therapy.Does not include description for genome editing.
	Considerations for quality and safety of gene therapy products using genome editing technology (Feb. 2020) [17].	Specific document for genome editing therapy.Detail description for considerations of quality and safety of in vivo and ex vivo genome editing.

EMA, European Medicines Agency; FDA, Food and Drug Administration; MHLW, Ministry of Health, Labour and Welfare; PMDA, Pharmaceuticals and Medical Devices Agency

EU EMA

The European Medicines Agency (EMA) has guidelines for in vivo gene therapy products and for ex vivo genetically modified cell products. The former guidelines, ‘Guideline on the quality, non-clinical and clinical aspects of gene therapy medicinal products’ [11] have been revised to take genome editing into account, though there is no specific description for genome editing.

On the other hand, the proposed amendment by EMA to the latter guideline, ‘Guideline on quality, non-clinical and clinical aspects of medicinal products containing genetically modified cells’ was necessitated partly by the need to address genome editing [12]. It details the quality and non-clinical considerations specific to genetically engineered cell products using genome editing. For instance, with respect to the characterization and non-clinical safety of genome-edited cells, it has been proposed that the genome-edited status of on-target and off-target sites and the ratio of genome-edited cells should be clarified. On-target sites should be examined to see how well they have been edited and whether any off-target alterations have occurred. The off-target sites should include not only those predicted by appropriate in silico analysis, but also those detected by in vitro analysis using cells, since the latter may include sites left undetected by in silico analysis. The proposed amendment also touches upon the types of genome editing tools and delivery methods used to generate genome edited cells, considerations regarding the persistence of genome editing enzymes in cells, and the selection of animal models for toxicity studies. This guideline is helpful in conducting the safety assessment of ex vivo genome-edited cell products.

US FDA

The US Food and Drug Administration (FDA) has also issued two revised guidance documents for gene therapy that address genome editing. An update of the guidance ‘Chemistry, Manufacturing, and Control (CMC) Information for Human Gene Therapy Investigational New Drug Applications (INDs)’ [13] addresses the use of genome editing tools, however, it does not touch upon the safety issues.

On the other hand, in the guidance ‘Long Term Follow-Up After Administration of Human Gene Therapy Products’ [14], the FDA states that genome-editing technologies may pose unknown and unpredictable risks to patients and may result in delayed adverse events. This could be attributed to the permanent alteration of the genome and the possibility that unintended genomic alterations may result in abnormal gene expression, chromosomal translocations, and induction of malignancies. For this reason, the study calls for a 15-year follow-up of patients who have undergone genome-edited gene therapy or gene therapy using integrating vectors, such as retroviral or lentiviral vectors, that are at risk for insertional mutagenesis. In addition, it calls for the development of a monitoring plan for late adverse events. The plan should be based on the results of in silico and in vitro analyses of the off-target effects of genome editing in non-clinical studies, which could determine the ratio of off-target to on-target editing expression and predict the off-target activity, based on the value of on-target activity. While the FDA does not have a guidance specific to the safety assessment of genome editing, this guidance provides specific instructions on the factors to be considered in the non-clinical safety studies of genome editing during patient follow-up.

Japan MHLW/PMDA

In Japan, the Ministry of Health, Labour and Welfare (MHLW) amended two guidelines for human gene therapy in 2019. In ‘Guidelines for Gene Therapy Clinical Research’ [15], genome editing therapy is defined as gene therapy. However, there are no specific guidelines regarding safety assessment of genome editing. In addition, ‘Guideline on Ensuring the Quality and Safety of Gene Therapy Products’ [16] refers only to conventional gene therapy, and does not include genome editing technologies.

Under these circumstances, the Science Board of the Pharmaceuticals and Medical Devices Agency (PMDA) discussed the quality and safety issues of genome editing gene therapy, based on published literature and the regulations in EU and US, and published a white paper, ‘Considerations for quality and safety of gene therapy products using genome editing technology’ [17] in Feb 2020. This white paper addressed the classification of genome editing technologies and their quality characteristics, the concept of safety assessment, and the points to be considered in clinical trials. Though it is not a guideline of MHLW, it is a useful reference for the industry and the reviewers of genome editing products in Japan.

Apart from the Science Board of PMDA, the AMED RS research group for genome editing carried out studies for the development of a draft technical guidance focusing on the safety assessment of unintended mutation of ex vivo genome-edited cell products [18]. This was done by discussions with the Genome Editing Therapy Safety Task Force of the Japan Pharmaceutical Manufacturers Association (JPMA), aided by data obtained from our experimental studies on CRISPR/Cas9 genome editing of human induced pluripotent stem (iPS) cells, published literature, and FDA and EMA guidelines. The following is a summary of the methods to evaluate off-target sites and the proposed technical guidance for evaluating unintended mutations of ex vivo genome-edited cells.

Methods to Evaluate Off-target Genome editing

Various methods have been developed to search for candidate off-target mutation sites by genome editing. These can be classified into in silico and in vitro analysis, with the latter divided into two categories-‘cell-based’ analysis, which uses whole cells, and ‘cell-free’ analysis, which utilizes the genome extracted from the cells (Table 2). Every method has its own disadvantages and requires a combination of several different methods of classification. A limited number of methods, including whole-genome sequencing (WGS) and amplicon deep sequencing, are available for the identification of off-target mutations that are occurring in genome-edited cell products, and these must be used to confirm the presence of mutations at candidate off-target sites.

Table 2. Comparison of methods to detect off-target sites

Category	Example [Ref.]	Method outline	Disadvantage
Search methods of candidate off-target sites
In silico analysis	GGGenome [1]CRISPRdirect [2]Cas-OFFinder [19]etc.	Prediction of sequence similarity to sgRNA using sequence search software.	May not predict every off-target site.
In vitro cell-free analysis	Digenome-seq [20, 21]SITE-seq [22]CIRCLE-seq [23]etc.	Genome-wide methods to search candidate off-target sites by enzymatic cleavage of genomic DNA extracted from the cells using genome editing components.	Does not necessarily reflect genome editing in the cell. Potential to overestimate.
In vitro cell-based analysis	GUIDE-seq [24]DISCOVER-seq [25]HTGTS [26, 27]etc.	Genome-wide methods to search candidate off-target cleavage sites by capturing tags etc. during genome editing of cells.	The efficiency of uptake into the cell and into the genome is affected.Potential to underestimate.
Identification methods of off-target sites
Cell-based analysis	Whole genome sequencing (WGS) using next-generation sequencing (NGS)	Comparison of WGS of cells before and after genome editing.	Difficult to distinguish between off-target mutations and spontaneous mutations in cell culture.
	Amplicon sequencing	PCR amplification and deep sequencing of the predicted/candidate cleavage sites.	Labor-intensive, in case the number of candidate sites are high.
	Digenome-seq [20, 21]	Comparison of Digenome-seq data of cells before and after genome editing.	Requires a large sample size and read depth.

In silico analysis

A common method for the in silico prediction of off-target candidate sites is to search the entire genome using a nucleotide sequence search program. However, since the CRISPR/Cas9 system recognizes only about 20 nucleotides in length, a common sequence search program may fail to recognize the sequence. Hence, it is important to use softwares such as GGGenome [1], CRISPRdirect [2] or Cas-OFFinder [19], which allow multiple mismatches and INDELs to be retrieved, even for such short sequences. Since it is difficult to accurately predict off-target mutations that occur in genome-edited cells using in silico analysis alone, the results of in vitro analysis too should be considered while predicting candidate off-target sites, as described below.

In vitro analysis

Cell-free analysis:

Cell-free analyses include methods to comprehensively detect off-target cleavage sites by enzymatic cleavage of genomic DNA extracted from cells with Cas9/sgRNA complex. These include Digenome-seq (in vitro Cas9-digested whole-genome sequencing) [20, 21], SITE-seq (selective enrichment and identification of tagged genomic DNA ends by sequencing) [22], and CIRCLE-seq (circularization for in vitro reporting of cleavage effects by sequencing) [23].

Digenome-seq is a method to comprehensively analyze the cleavage sites by applying Cas9/sgRNA complex to genomic DNA, followed by analysis of the fragments obtained by WGS using next-generation sequencing (NGS). In addition to using the genomes of unedited cells to detect off-target sites, the system can also be used to evaluate the presence of off-target mutations in genome-edited cells and to evaluate the efficiency of genome editing of on- and off-target sites. This is possible due to the fact that the sequence is altered by genome editing and is no longer cleaved by Cas9/sgRNA complex. The disadvantage of this method is its cost, since it requires high-depth WGS.

On the other hand, SITE-seq is a method to detect the cleavage site by the addition of a biotin-modified adapter to the cleavage site by Cas9/sgRNA, enrichment of the modified sequence with streptavidin and the detection of the cleavage site by NGS analysis. The CIRCLE-seq is also a method for detecting the cleavage site by NGS analysis and involves the enrichment of the straight stranded sequences alone, which are obtained after cyclization of mechanically fragmented genomic DNA and cleavage with Cas9/sgRNA. Both SITE-seq and CIRCLE-seq can perform NGS analysis at a lower cost compared to Digenome-seq, since the off-target cleavage sites are enriched for NGS analysis. However, unlike Digenome-seq, it does not directly detect the presence of off-target mutations in genome-edited cells.

The cell-free analysis has the potential to detect candidate off-target cleavage sites with better sensitivity than cell-based analysis, since it is unaffected by the efficiency of the genome editing system’s introduction into cells or the labeling of the genome. However, the structure of genomic DNA extracted from cells is different from their structure in the cell and does not necessarily reflect DNA cleavage by genome editing in the cell. In addition, cell-free analysis requires the genome editing enzyme itself, which cannot be directly evaluated in systems that express the CRISPR/Cas9 in cells using virus/plasmid vectors or mRNAs.

Cell-based analysis:

Cell-based analyses are genome-wide unbiased methods to screen candidate off-target sites by enriching DSB sites during genome editing in cells. These methods include GUIDE-seq (genome-wide, unbiased identification of DSBs enabled by sequencing) [24], which utilizes the incorporation of double-stranded DNA (dsDNA) into DSB; DISCOVER-seq (discovery of in situ Cas off-targets and verification by sequencing) [25], which utilizes the binding of MRE11 proteins to DSB sites for DSB repair; and HTGTS (high-throughput genome-wide translocation sequencing) [26, 27], which utilizes translocation as an indicator for detection. These methods may not be able to detect all candidate off-target sites with high sensitivity, given the impact of the genome editing system’s efficiency in transferring the genome to the cells and the labeling efficiency of the genome. In addition, it does not directly detect mutations in genome-edited cell products. On the other hand, the expression of genome editing systems in vectors or mRNAs may also be evaluated using these methods.

A more direct method for detecting off-target mutations in edited cell products is WGS. Comparison of WGS of cells before and after genome editing should, in principle, comprehensively detect all off-target mutations caused by genome editing. In practice, however, it is difficult to directly compare the whole-genome sequence data from cells before and after genome editing, and hence, the obtained sequences are mapped to the reference genome for comparison. In addition, since natural mutations are introduced in the genome during the cell culture process, it is difficult to distinguish whether the mutations detected are off-target mutations due to genome editing, natural mutations, sequencing errors or mutations due to other causes. For the direct evaluation of genome-edited cell products, it is necessary to evaluate the mutation status at the candidate off-target sites narrowed down on by combining the in silico analysis and in vitro cell-free or cell-based analysis. In this case, amplicon sequencing can be used in addition to WGS and Digenome-seq. Amplicon sequencing is a method of NGS analysis which involves the amplification of only the narrowed-down candidate off-target sites. While it is possible to assess the status of mutations at the predicted site sensitively with fewer reads than WGS, it is impossible to detect unpredicted mutations occurring in the cell. In addition, if there are a large number of candidate sequences, it becomes necessary to perform PCR for a very large number of combinations.

Proposed Technical Guidance on the Methods of Safety Assessment for Unintended Genetic Modification of Genome-Edited Cell Products

Genome editing may result in unintended genetic modification, inducing cell growth arrest, cell death, and changes in cellular properties. The genome-edited cell products pose a significant risk of changes in cell properties, especially tumorigenesis of cells [14], necessitating the evaluation of these risks for genome-edited cell products. The following is the summary of a proposed technical guidance for evaluating unintended mutations of ex vivo genome-edited cells developed by our AMED RS research project for genome editing [18].

Strategy to assess off-target mutations of ex vivo genome-edited cell products

Genome-edited cell products are classified into two types- cloned cells and non-cloned cells after genome editing, with each requiring a different strategy for the assessment of off-target mutation (Fig.1). In the case of cloned cells, the genome editing state is uniform across cells and allows for assessment of off-target mutations of the genome-edited cell products themselves. However, in the case of non-cloned cells, the genome edited state is diverse and even if a few cells are evaluated, it is not possible to capture the whole picture of off-target editing. In this case, the possible off-target mutations caused by the genome editing methods used should be evaluated rather than the genome-edited cell product itself.

Fig. 1.

Strategy to assess off-target mutations of ex vivo genome-edited cell products.

Cloned cells:

In pluripotent stem cells such as iPS cells, only the target genome-edited cells are cloned after genome editing, and once banked, they are differentiated into the target therapeutic cells and administered to the patient. In this case, it is advisable to have the banked source cells or the final product evaluated by both in silico and in vitro analysis [12], according to the procedure detailed below.

1. Predict the candidate off-target sites by in silico analysis (ex. GGGenome [1], CRISPRdirect [2], Cas-OFFinder [19]).

2. Search for candidate off-target cleavage sites by in vitro analysis (cell-free analysis such as Digenome-seq [20, 21], SITE-seq [22], CIRCLE-seq [23] or cell-based analysis such as GUIDE-seq [24], DISCOVER-seq [25], HTGTS [26, 27]).

3. Confirm the presence or absence of off-target mutations at candidate sites obtained in steps 1 and 2 in genome-edited cloned cells (final product or banked source cells) by WGS, amplicon sequencing, or Digenome-seq [20, 21].

4. If off-target mutations are found in genome-edited cells, the location and type of the mutation is confirmed, followed by a risk assessment of the mutated gene based on the presence or absence of mutations in cancer-related genes, such as p53 [9, 17].

Non-cloned cells:

T cells and hematopoietic stem cells are difficult to clone after genome editing, hence it is assumed that they will be administered to patients as pooled cells. In this case, it is difficult to perform WGS because of the presence of a mixture of cells with different genome editing states. Since it is not possible to evaluate the genome-editing cell product itself for therapeutic use in non-cloned cells, it is desirable to evaluate the potential off-target effects of the genome-editing method following the procedure detailed below.

1. Predict candidate off-target sites by in silico analysis [1, 2, 19].

2. Search for candidate off-target cleavage sites by in vitro analysis (cell-free or cell-based analysis [20,21,22,23,24,25,26,27]).

3. Perform genome editing using the same type of cells as the target cells by the genome editing system. The presence or absence of off-target mutations at candidate sites obtained in steps 1 and 2 is confirmed in genome-edited cells by amplicon sequencing or Digenome-seq [20, 21].

4. The proportion of genome-edited cells of interest and cell heterogeneity, including the editing efficiency of on- and off-target sites, should be clarified [12, 14].

5. If off-target mutations are found in genome-edited cells, the location and type of the mutation is confirmed, followed by a risk assessment of the mutated gene based on the presence or absence of mutations in cancer-related genes, such as p53 [9, 17].

Evaluation of unintended sequence insertions, deletions and inversions at on-target site

The safety assessment of the on-target site requires an appropriate method to assess the possibility of large sequence insertions/large deletions and chromosomal translocations [4, 5]. In the case of cloned cells after genome editing, the insertion of unintentional sequences can be detected by PCR amplification with primers positioned between the on-target sites and electrophoresis of the amplified fragments. However, in the case of significant deletions or inversions, the position of the primer setting should be considered [5]. If insertions, deletions, or inversions are identified, a risk assessment is performed.

In cells that do not clone after genome editing, if the non-intentional sequence inserted at the on-target site is up to several hundred bases, the genomic region containing the target site of the genome-edited cell can be amplified by PCR and evaluated by Amplicon sequencing with short-read NGS. For longer insertions, if the sequence containing the target site can be amplified by PCR, an exhaustive analysis using a third-generation sequencer is also possible [5]. In the case of large deletions or inversions, as in the case of cloned cells, the position of the primers should be considered [5, 17]. In addition, inversions and translocations may occur, making PCR primer design difficult. Thus, it is necessary to check the nucleotide sequence and chromosome changes of genome-editing cells for the possibility of unintended genetic modifications. If insertions, deletions, or inversions are found, risk assessment is done considering their frequency and reproducibility.

It should be noted that these mutations have already been reported to occur not only at on-target sites, but also at off-target sites [6,7,8].

Evaluation of chromosomal abnormalities

The analysis of chromosome aberrations caused by translocations cannot be done by NGS analysis. The methods for evaluating chromosome aberrations include G- and Q-band analysis, multicolor fluorescence in situ hybridization (mFISH) using pseudo colors, and comparative genomic hybridization (CGH) using arrays. However, it is important to note that there are certain limitations to its detection [17]. G-band analysis and mFISH can only analyze cells during metaphase. In addition, G-band analysis is not suitable for detecting chromosomal abnormalities in minor cell populations, since it is difficult to observe a large number of cells. While mFISH can detect translocations between different chromosomes and large deletions within a chromosome, it cannot detect inversions within the same chromosome. CGH can detect amplification or deletion of abnormal genes in many cells, but it is not sensitive enough to detect heterogeneity between cells or abnormalities that occur in some cells.

Evaluation of genome-edited cell products for tumorigenicity

Particular attention should be paid to cell carcinogenesis while assessing the risk of unintended genetic modification by genome editing. Depending on the characteristics of the cells used and the corresponding diseases, it is necessary to examine the changes in the proliferative properties of the cells and the need for evaluation of tumorigenicity.

The risk of tumorigenesis of genome-edited cells can vary depending on the type and nature of the cells (cell longevity, undifferentiated nature). Pluripotent stem cells, such as iPS/ES cells, are considered to be the most undifferentiated and associated with high risk. They are followed by somatic stem cells, such as hematopoietic stem cells, while differentiated somatic cells, such as T cells, are considered to be low risk despite their proliferative nature, based on previous experiences with gene therapy [17].

Conclusion

In this article, the regulatory and safety issues of genome editing technologies for human gene therapy are summarized, and the proposed technical guidance for the safety assessment of ex vivo genome-edited cell products developed by the AMED RS research group for genome editing is introduced. The concept of the strategy for safety assessment of off-target editing presented here is almost similar to that of the EU [12] and US [14] guidelines and the report of PMDA Science Board in Japan [17]. As for the safety assessment of in vivo genome editing, since the evaluation is difficult in animals owing to the difference of genomes between animals and humans, it is advisable to perform the evaluation primarily in human cells. Further discussion is needed on certain aspects, such as the extent to which the assessment in cells reflects the in vivo genome editing, the type of cells to be used and the method of their assessment. It is hoped that this guidance will promote the development of genome editing-based gene therapy in Japan.

Funding

This work was supported by the Japan Agency for Medical Research and Development (AMED).

Acknowledgements

The author thanks the members of the AMED RS research project for genome editing (Dr. Yuki Naito, Dr. Ryuichi Ono, Dr. Takuma Yamashita, Dr. Takao Inoue, and Dr. Teruhide Yamaguchi) and the Genome Editing Therapy Safety Task Force of JPMA for fruitful discussion.

References

• 1.  GGGenome.https://GGGenome.dbcls.jp/ (Naito Y. 2014. GGGenome. Experimental Medicine32: 3263–3264.; Yoshida T, Naito Y, Sasaki K, Uchida E, Sato Y, Naito M, Kawanishi T, Obika S, Inoue T. 2018. Estimated number of off-target candidate sites for antisense oligonucleotides in human mRNA sequences. Genes Cells23, 448–455) [accessed May 1, 2020].
• 2.  CRISPRdirect.https://crispr.dbcls.jp/ (Naito, Y., Hino, K., Bono, H., Ui-Tei, K. 2015. CRISPRdirect: Software for designing CRISPR/Cas guide RNA with reduced off-target sites. Bioinformatics31: 1120–1123) [accessed May 1, 2020].
• 3.   Wang,  H.,  Yang,  H.,  Shivalila,  C. S.,  Dawlaty,  M. M.,  Cheng,  A. W.,  Zhang,  F. and  Jaenisch,  R. 2013. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell 153: 910–918.
• 4.   Ono,  R.,  Ishii,  M.,  Fujihara,  Y.,  Kitazawa,  M.,  Usami,  T.,  Kaneko-Ishino,  T.,  Kanno,  J.,  Ikawa,  M. and  Ishino,  F. 2015. Double strand break repair by capture of retrotransposon sequences and reverse-transcribed spliced mRNA sequences in mouse zygotes. Sci. Rep. 5: 12281.
• 5.   Kosicki,  M.,  Tomberg,  K. and  Bradley,  A. 2018. Repair of double-strand breaks induced by CRISPR-Cas9 leads to large deletions and complex rearrangements. Nat. Biotechnol. 36: 765–771.
• 6.   Moore,  J. K. and  Haber,  J. E. 1996. Capture of retrotransposon DNA at the sites of chromosomal double-strand breaks. Nature 383: 644–646.
• 7.   Bill,  C. A. and  Summers,  J. 2004. Genomic DNA double-strand breaks are targets for hepadnaviral DNA integration. Proc. Natl. Acad. Sci. USA 101: 11135–11140.
• 8.   Onozawa,  M.,  Zhang,  Z.,  Kim,  Y. J.,  Goldberg,  L.,  Varga,  T.,  Bergsagel,  P. L.,  Kuehl,  W. M. and  Aplan,  P. D. 2014. Repair of DNA double-strand breaks by templated nucleotide sequence insertions derived from distant regions of the genome. Proc. Natl. Acad. Sci. USA 111: 7729–7734.
• 9.   Haapaniemi,  E.,  Botla,  S.,  Persson,  J.,  Schmierer,  B. and  Taipale,  J. 2018. CRISPR-Cas9 genome editing induces a p53-mediated DNA damage response. Nat. Med. 24: 927–930.
• 10.   Ihry,  R. J.,  Worringer,  K. A.,  Salick,  M. R.,  Frias,  E.,  Ho,  D.,  Theriault,  K.,  Kommineni,  S.,  Chen,  J.,  Sondey,  M.,  Ye,  C.,  Randhawa,  R.,  Kulkarni,  T.,  Yang,  Z.,  McAllister,  G.,  Russ,  C.,  Reece-Hoyes,  J.,  Forrester,  W.,  Hoffman,  G. R.,  Dolmetsch,  R. and  Kaykas,  A. 2018. p53 inhibits CRISPR-Cas9 engineering in human pluripotent stem cells. Nat. Med. 24: 939–946.
• 11.  European Medicines Agency. Guideline on the quality, non-clinical and clinical aspects of gene therapy medicinal products, March 22, 2018. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-quality-non-clinical-clinical-aspects-gene-therapy-medicinal-products_en.pdf [accessed May 1, 2020].
• 12.  European Medicines Agency. Draft guideline on quality, non-clinical and clinical aspects of medicinal products containing genetically modified cells, July 26, 2018. https://www.ema.europa.eu/en/documents/scientific-guideline/draft-guideline-quality-non-clinical-clinical-aspects-medicinal-products-containing-genetically_en.pdf [accessed May 1, 2020].
• 13.  Food and Drug Administration. Chemistry, Manufacturing, and Control (CMC) Information for Human Gene Therapy Investigational New Drug Applications (INDs), Guidance for Industry, January, 2020. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/chemistry-manufacturing-and-control-cmc-information-human-gene-therapy-investigational-new-drug [accessed May 1, 2020].
• 14.  Food and Drug Administration. Long Term Follow-Up After Administration of Human Gene Therapy Products, Guidance for Industry, January, 2020. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/long-term-follow-after-administration-human-gene-therapy-products [accessed May 1, 2020].
• 15.  Ministry of Health Labour and Welfare, Japan. Guidelines for gene therapy clinical research, February 28, 2019 (in Japanese). https://www.mhlw.go.jp/content/000561788.pdf [accessed May 1, 2020].
• 16.  Ministry of Health Labour and Welfare, Japan. Guideline on Ensuring the Quality and Safety of Gene Therapy Products, July 9, 2019 (in Japanese). https://www.pmda.go.jp/files/000230508.pdf [accessed May 1, 2020].
• 17.  Science Board, Pharmaceuticals and Medical Devices Agency. White paper: Considerations for quality and safety of gene therapy products using genome editing technology, February 7, 2020. https://www.pmda.go.jp/files/000233744.pdf (in Japanese) [accessed May 1, 2020].
• 18.   Uchida,  E.,  Yamashita,  T.,  Ono,  R.,  Naito,  Y., Task Force for Safety Assessment of Genome Editing Therapy, Inoue, T. 2019. Regulations and Safety evaluation for genome editing therapy. Pharmaceutical and Medical Device Regulatory Science 50: 513–522 (in Japanese).
• 19.  Cas-OFFinder.http://www.rgenome.net/cas-offinder/ (Bae, S., Park J., Kim, J.S. 2014. Cas-OFFinder: a fast and versatile algorithm that searches for potential off-target sites of Cas9 RNA-guided endonucleases. Bioinformatics30: 1473–1475 [accessed May 1, 2020].
• 20.   Kim,  D.,  Kim,  S.,  Kim,  S.,  Park,  J. and  Kim,  J. S. 2016. Genome-wide target specificities of CRISPR-Cas9 nucleases revealed by multiplex Digenome-seq. Genome Res. 26: 406–415.
• 21.   Kim,  D.,  Bae,  S.,  Park,  J.,  Kim,  E.,  Kim,  S.,  Yu,  H. R.,  Hwang,  J.,  Kim,  J. I. and  Kim,  J. S. 2015. Digenome-seq: genome-wide profiling of CRISPR-Cas9 off-target effects in human cells. Nat. Methods 12: 237–243, 1, 243.
• 22.   Cameron,  P.,  Fuller,  C. K.,  Donohoue,  P. D.,  Jones,  B. N.,  Thompson,  M. S.,  Carter,  M. M.,  Gradia,  S.,  Vidal,  B.,  Garner,  E.,  Slorach,  E. M.,  Lau,  E.,  Banh,  L. M.,  Lied,  A. M.,  Edwards,  L. S.,  Settle,  A. H.,  Capurso,  D.,  Llaca,  V.,  Deschamps,  S.,  Cigan,  M.,  Young,  J. K. and  May,  A. P. 2017. Mapping the genomic landscape of CRISPR-Cas9 cleavage. Nat. Methods 14: 600–606.
• 23.   Tsai,  S. Q.,  Nguyen,  N. T.,  Malagon-Lopez,  J.,  Topkar,  V. V.,  Aryee,  M. J. and  Joung,  J. K. 2017. CIRCLE-seq: a highly sensitive in vitro screen for genome-wide CRISPR-Cas9 nuclease off-targets. Nat. Methods 14: 607–614.
• 24.   Tsai,  S. Q.,  Zheng,  Z.,  Nguyen,  N. T.,  Liebers,  M.,  Topkar,  V. V.,  Thapar,  V.,  Wyvekens,  N.,  Khayter,  C.,  Iafrate,  A. J.,  Le,  L. P.,  Aryee,  M. J. and  Joung,  J. K. 2015. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases. Nat. Biotechnol. 33: 187–197.
• 25.   Wienert,  B.,  Wyman,  S. K.,  Richardson,  C. D.,  Yeh,  C. D.,  Akcakaya,  P.,  Porritt,  M. J.,  Morlock,  M.,  Vu,  J. T.,  Kazane,  K. R.,  Watry,  H. L.,  Judge,  L. M.,  Conklin,  B. R.,  Maresca,  M. and  Corn,  J. E. 2019. Unbiased detection of CRISPR off-targets in vivo using DISCOVER-Seq. Science 364: 286–289.
• 26.   Frock,  R. L.,  Hu,  J.,  Meyers,  R. M.,  Ho,  Y. J.,  Kii,  E. and  Alt,  F. W. 2015. Genome-wide detection of DNA double-stranded breaks induced by engineered nucleases. Nat. Biotechnol. 33: 179–186.
• 27.   Chiarle,  R.,  Zhang,  Y.,  Frock,  R. L.,  Lewis,  S. M.,  Molinie,  B.,  Ho,  Y. J.,  Myers,  D. R.,  Choi,  V. W.,  Compagno,  M.,  Malkin,  D. J.,  Neuberg,  D.,  Monti,  S.,  Giallourakis,  C. C.,  Gostissa,  M. and  Alt,  F. W. 2011. Genome-wide translocation sequencing reveals mechanisms of chromosome breaks and rearrangements in B cells. Cell 147: 107–119.