Off-target effects and overall efficacy are major considerations in the design of tools for genome editing. In a recent issue of Trends in Biotechnology, Hendel et al review methods used to assess both the efficacy of genome editing technologies as well as methods to limit off-target effects.
In the first half of their review, Hendel et al. discuss methods of qualifying and quantifying genome editing events.
The analysis of genome editing is broken down into mutagenic non-homologous end joining (NHEJ) events and donor DNA-driven homologous recombination (HR) events. NHEJ events can be identified using high resolution melt curve analysis, or through gel-based analysis following CEL-1 or T7 endonuclease treatment (Fig. 1a). A similar PCR/gel based approach can be used to identify HR events through the incorporation of a unique restriction site within the donor (Fig. 1b). Hendel et al. also point out that direct Sanger sequencing can be used, as well as high-throughput next generation sequencing, which is capable of detecting NHEJ events at frequencies as low as 0.01%.
These PCR based technologies, however, are qualitative and do not provide a robust estimate of the genome editing activity.
To address this, Hendel et al. review several technologies used to quantify the activity of genome editing technologies (where they consider 25% activity to represent “highly active” nucleases). One means of quantification is through the use of a traffic light reporter (TLR). In TLR, two fluorescent proteins are present with the second protein being out of frame, while the first fluorescent protein is in frame, but non-functional (Fig 2). If a NHEJ event occurs, the second protein is restored to in-frame and is expressed. If an HR event occurs, the first protein will, instead, be replaced by a functional copy, and expressed. The treated cells are then quantified using flow cytometry. While the TLR assay is ideal for quantifying NHEJ and HR events, it is limited to cell culture and highly-precise genome editing events.
A more robust approach, that is independent of the requirements of TLR, is the use of next generation sequencing. Illumina and 454 sequencing can detect NHEJ and HR events, but do not allow for long reads to sequence entire HR events. Recently, Hendel et al., modified single molecule real time (SMRT) sequencing to allow for long reads that can span HR events (Fig 3). This allows for not only the quantification of NHEJ and HR events, but also provides in depth information of the various DNA repairs that occurred.
In the second half of their review, Hendel et al. discuss means of quantifying off-target events in genome editing, which they separate into in vitro, cellular, and in silico analyses.
Perhaps the most elegant of the methods outlined, is the systematic evolution of ligands by exponential enrichment (SELEX), in which a semi-randomized library of oligonucleotides possessing constant 5’ and 3’ end sequences, is exposed to the genome editing technology in vitro, and the surviving fragments can be amplified through PCR allowing for the identification of off-targets.
While the review is largely aimed at summarizing the details of genome editing for clinical use, many of the mentioned technologies within this review are applicable to insect systems.
Overall, the review can serve as a good starting point to identify the various tools (from low-tech to high-tech) that can be used to identify on- and off-targets in genome editing.