Preclinical safety assays to detect and define the genotoxic effects of nucleases and homology donors
| Class of assay . | Methods . | Advantages . | Disadvantages . |
|---|---|---|---|
| Sequence based | Genome-wide sequencing | Unbiased assessment of entire genome | Cannot be used on populations of cells because of sensitivity |
| Background error rate remains high | |||
| Bioinformatics based: use computational tools to predict off-target sites and then deep sequence predicted sites to assess whether insertions/deletions are being created | Many currently available algorithms and sequencing platforms make this easily available | Prediction algorithms have not routinely identified all of the off-target sites that have been found using other methods | |
| Karyotyping | Routine procedure | Only assesses a small number of cells (20-50) | |
| Unbiased DSB capture methods combined with sequencing; can be done by capturing break with plasmid DNA,21 AAV,22 IDLV,23,24 or oligos,25 or by directly sequencing sites of DSBs (Digenome Seq26 or BLESS27 ). | Unbiased | Technically difficult with sensitivity of 0.1% | |
| Translocation capture28 | Unbiased assessment for genome rearrangements | Technically difficult; primarily assesses rearrangements from the intended target site not between 2 off-target sites | |
| “Onco-chip” | Specifically assesses for mutations in genes associated with cancer | Low sensitivity; list may be incomplete | |
| Function based | Staining for DSBs (53BP129 or gH2AX30 ) | Simple and quantitative and can be applied to a wide variety of cell types and used in vivo | No information on sites of DSBs or consequence of DSBs |
| Relative cell proliferation29 | Simple and quantitative and can be applied to a wide variety of cell types | Does not give information on mechanisms | |
| Cell cycle perturbations30 | Quantitative measure of functional effect on key cell function | Only used in a specialized cancer cell line | |
| Clonal dynamics assays31 | Can assess whether nucleases can induce changes in clonal representation in large populations of cells | Requires cell population to be grown for long periods of time; could be confounded by the integration site of barcode to mark clones | |
| Colony-forming replating assay32 | Useful assessment for retroviral and lentiviral vectors | Not clinically validated and has not yet been adopted to genome-editing strategies | |
| Lineage reconstitution and colony-forming assays5 | Direct functional measure of edited cell’s behavior | Sensitivity to important genotoxic events is not validated | |
| In vitro transformation33 | The best current model for assessing cell transformation | Very insensitive and not validated for predicting human |
| Class of assay . | Methods . | Advantages . | Disadvantages . |
|---|---|---|---|
| Sequence based | Genome-wide sequencing | Unbiased assessment of entire genome | Cannot be used on populations of cells because of sensitivity |
| Background error rate remains high | |||
| Bioinformatics based: use computational tools to predict off-target sites and then deep sequence predicted sites to assess whether insertions/deletions are being created | Many currently available algorithms and sequencing platforms make this easily available | Prediction algorithms have not routinely identified all of the off-target sites that have been found using other methods | |
| Karyotyping | Routine procedure | Only assesses a small number of cells (20-50) | |
| Unbiased DSB capture methods combined with sequencing; can be done by capturing break with plasmid DNA,21 AAV,22 IDLV,23,24 or oligos,25 or by directly sequencing sites of DSBs (Digenome Seq26 or BLESS27 ). | Unbiased | Technically difficult with sensitivity of 0.1% | |
| Translocation capture28 | Unbiased assessment for genome rearrangements | Technically difficult; primarily assesses rearrangements from the intended target site not between 2 off-target sites | |
| “Onco-chip” | Specifically assesses for mutations in genes associated with cancer | Low sensitivity; list may be incomplete | |
| Function based | Staining for DSBs (53BP129 or gH2AX30 ) | Simple and quantitative and can be applied to a wide variety of cell types and used in vivo | No information on sites of DSBs or consequence of DSBs |
| Relative cell proliferation29 | Simple and quantitative and can be applied to a wide variety of cell types | Does not give information on mechanisms | |
| Cell cycle perturbations30 | Quantitative measure of functional effect on key cell function | Only used in a specialized cancer cell line | |
| Clonal dynamics assays31 | Can assess whether nucleases can induce changes in clonal representation in large populations of cells | Requires cell population to be grown for long periods of time; could be confounded by the integration site of barcode to mark clones | |
| Colony-forming replating assay32 | Useful assessment for retroviral and lentiviral vectors | Not clinically validated and has not yet been adopted to genome-editing strategies | |
| Lineage reconstitution and colony-forming assays5 | Direct functional measure of edited cell’s behavior | Sensitivity to important genotoxic events is not validated | |
| In vitro transformation33 | The best current model for assessing cell transformation | Very insensitive and not validated for predicting human |
BLESS, breaks labeling, enrichment on streptavidin, and next-generation sequencing; IDLV, integration-deficient lentiviral vector; oligo, oligonucleotide.