The development of efficient and precise genome editing tools like CRISPR/Cas9 call for improved methods for fast, precise and in-depth dissection of the insertion or deletion (indels) of bases in the DNA sequence. The traditional methodologies include enzyme mismatch cleavage assay (EMC) and DNA sequencing. EMC lacks sensitivity, resolution and high-throughput capability, and the sequencing of DNA is expensive and time-consuming. The development of the indel detection by amplicon analysis (IDAA) methodology may meet the requirements for fast, precise, sensitive and cost effective indel profiling in genome editing applications.

Gene editing tools

The nuclease-based gene editing tools such as meganucleases, zinc-finger nucleases (ZFN), transcription activator-like effector nucleases (TALEN), and CRISPR/Cas9 are rapidly transforming capabilities for altering the genome of cells and organisms with great precision and in high throughput studies. In Denmark, we were among the first to embark on these technologies in mammalian cells. We used ZFN gene targeting in detailed analysis of glycoproteomes to resolve the extensive heterogeneities in glycan structures and attachment sites. ZFN targeting has been applied to truncate the O-glycan elongation pathways in human cells, generating stable ‘SimpleCell’ lines with homogenous O-glycosylation (1).

The introduction of CRISPR/Cas9 gene editing is a breakthrough, which has great potential in basic and biomedical research. Clearly, the simplicity and general access to CRISPR/Cas9 reagents has democratized the conduct of precise genome targeting in cells, tissues, organs, and whole animals.

Figure 1. CRISPR/Cas9 gene editing. Workflow from CRISPR/Cas9 cell delivery, DNA extraction, and indel profiling to insight. Notably, the time in days from CRISPR/Cas9 delivery to DNA extraction normally lasts for 2-3 days, while time from indel analysis to insight depends on the indel detection methodology used and may take from 1 day to more than 5 days.

Lack of efficient evaluation methods

A major limitation in application of precise gene editing lies in the lack of sensitive and fast methods to detect and characterize the DNA changes introduced by CRISPR/Cas9. Precise gene editing induces double-stranded DNA breaks that are repaired by non-homologous end joining, which is error-prone, leading to introduction of indels at the target site. These indels are often small, difficult, and laborious to detect by traditional methods (2).  This limits screening, detection and selection of the desired targeting events in cells, tissues and whole animals. Furthermore, the detection of indels induced by CRISPR/Cas9 should be accomplished by an automated high-throughput methodology that easily integrates into routine genome editing workflows. Ideally, the time required for indel profiling of a CRISPR/Cas9 targeted DNA sample should be as short as possible (Figure 1).

Current methods for accurate identification of the genetic modifications are few, laborious and require extensive screening of clones in order to identify correctly edited bi-, tri- or multi allelic genes. Moreover, the indel detection methods including NGS, Sanger sequencing and EMC lack sufficient speed and sensitivity (Figure 2).

In order to meet this challenge we have developed methods and protocols for fast, sensitive and cost-efficient detection of indels induced by precise genome targeting (2-4). Through the Consortium for Designer Organisms at the University of Copenhagen (http://www.cdo.ku.dk/), these methods have successfully been made available to national and international academic and biopharmaceutical groups working in the field of genome targeting.

Figure 2. Indel detection analysis in genome targeting. Assay time and performance for four indel detection methods: NGS, Sanger sequencing, EMC and IDAA. Time in days on axis shown below for the required steps (blue hexagons) for each method  from genome targeted cell pool (grey hexagon) to insight. QC (white triangle) represents necessary quality control steps. The table shown to the right displays the performance (high or low) of the respective methods with regard to sensitivity, resolution, high through-put amenability (HTRP) and cost. Preferred choices are shown in bold font.

Indel detection methodologies

The choice of indel detection methodology to be used is primarily determined by the scope of the experiment and available analytical resources. Currently, only a few methods enable assessment of the full spectrum of repair products in a cell population following CRISPR/Cas9 cutting. Although evaluation of indel sequences by next generation sequencing (NGS) methods may be the best route to assess the full spectrum of CRISPR/Cas9 induced repair events, the time consumption and high cost constrain widespread adoption of this method. Therefore, the most commonly used indel detection method is the EMC assay (5).

Enzyme Mismatch Cleavage assay

EMC assays are based on selective endonuclease recognition and restriction/cleavage of heteroduplex DNA (not homoduplex DNA) formed after reannealing of heterogeneous amplicons possessing differentnucleotide variations, including indels (5). Thus the readouts of EMC assays are the cleaved amplicon fragments detected following endonuclease treatment of samples. EMC assays only require gel based electrophoresis and commercially available endonucleases such as T4 endonuclease VII (T4E7), endonuclease V (EndoV), T7 endonuclease I (T7EI), CELI or Surveyor nuclease.

However, the efficiency of EMC has been shown to depend on the properties of the mismatch-cleavage endonuclease. In particular T7EI and Surveyor possess a strong nonspecific activity and display a preference for heteroduplex DNA formed by deletions rather than single point mutations (4). Furthermore, CRISPR/Cas9 genome targeting predominantly generate minor indels (4,6), and endonuclease T7EI poorly discriminates the presence of single base deletion events (2). Therefore, EMC assays underestimates the editing efficiency in cells following CRISPR/Cas9 genome editing experiments.

DNA Sequencing

Optimal discrimination of DNA fragments down to single base differences has until recently only been possible by Sanger sequencing and NGS. By these approaches, target derived amplicons are cloned into plasmids or sequenced directly followed by bioinformatic deconvolution of the individual sequences and indels identified. Procedures such as tracking of indels by decomposition (TIDE) (https://tide.nki.nl/) have streamlined the workflow of amplicon cloning and sequencing. However, indel profiling is limited by indel size, lack of sensitivity and resolution for indel profiling of samples with high indel complexity (such as indels in cell pools) and is dependent on pairs of high quality Sanger sequencing data.

NGS is a powerful and widely accepted method for in-depth indel identification of a targeted genomic locus, but requires laborious, multi-step, skillful preparation of the amplicon libraries to be sequenced, and complex bioinformatics data analysis (Figure 2). Several online resources have been developed for indel quantification and characterization of NGS data, including CRISPR Genome Analyzer (http://54.80.152.219/), CRISPResso (http://crispresso.rocks/) and CRISP-R (https://bioconductor.org/packages/release/bioc/html/CrispRVariants.html), that all provide a complete reports containing all NGS information.

Indel detection by amplicon analysis methodology

To meet the requirement for fast, precise, cost-efficient indel detection with down to single base discrimination, we turned to a sequenator based application based on fragment analysis (FA) commonly applied to linkage, microsatellite instability and loss of heterozygosity mapping and SNP genotyping. We have recently shown that by combining FA with a simple amplicon labelling strategy for unbiased and simple homogeneous fluorophore labelling of amplicons, we enabled fast, sensitive and robust profiling of indels induced by CRISPR/Cas9 in cell clones, cell pools and whole animals (2-4). We showed in a “head to head” comparison of IDAA vs. NGS that the sensitivity and resolution of IDAA is comparable to NGS with an indel detection sensitivity ≈0.1% (4). Furthermore, IDAA is robust generating near identical profiles on independently repeated experiments. Finally, IDAA can be used for indel profiling regardless of the nature of the tools used to induce the indels analyzed, i.e. ZFNs, TALENs or CRISPR/Cas9.

An example of a typical CRISPR/Cas9 experiment analyzed by IDAA is shown in Figure 3. As can be seen, individual indels are clearly identified as single peaks within the chromatogram. Of notice, the most frequently occurring indel is a one base insertion, observed with 38% frequency. Rare deletion indels are found at lower frequencies. The total number of indels is estimated to 58% and thus, the CRISPR/Cas9 genome targeting efficiency (i.e. cutting efficiency) is 58%. Furthermore, 42% of the cellular targets (alleles) remain intact in size, which may be an acceptable level of un-targeted alleles depending on the aim of the genome targeting experiment conducted. Lastly, the 96-plate format of IDAA makes the assay ideal for high through-put genome targeting applications.

Figure 3. IDAA profile. The profile was generated from CRISPR/Cas9 targeting of the cosmc gene using our recently published protocol (4) Lipofectamine delivered CRISPR/Cas9 plasmids of CHO cells was followed by; 1) DNA extraction 2 days post transfection, 2) tri-primer fluorophore PCR amplification and 3) IDAA analysis. The turnaround time for full IDAA analysis from sample to insight is 5 hours, see Figure 2. Individual indels can be quantified, with individual indel frequencies shown at the peaks. A wild-type peak is marked by wt (red). Furthermore, the validated cutting efficiency of the CRISPR/Cas9 design used (58%) in this example is shown in bold in the upper left corner.

The chemistry of the assay is cost-efficient, requiring cheap and commercially available reagents in the form of primers available in a kit format (http://tagc.dk/). Downstream analysis requires commonly available DNA sequenator instrumentation, often not used for Sanger sequencing anymore, but stashed and hidden away in laboratory store rooms. Alternatively, the latter analytical part can be handled by custom service providers of IDAA analysis (https://www.eurofinsgenomics.eu/5746.aspx), by shipment of the fluorophore labelled amplicons for analysis and return of data profiles similar to the profile shown in Figure 3.Raw data files can easily be reanalyzed using freely available software, as described in ref. 4.

The recent development of highly efficient precise genome editing tools, such as ZFN’s, TALEN’s and CRISPR/Cas9, call for improved methods for fast and in-depth dissection of the changes in DNA sequence. EMC lacks sufficient sensitivity, resolution and high through-put capability and DNA sequencing is prohibitively expensive and time consuming for routine use in research laboratories. The development of the IDAA methodology may meet the requirements for fast, precise, sensitive and cost-effective indel profiling in basic research and more challenging genome editing applications such as therapeutic indel profiling.

References

1. Steentoft, C., Vakhrushev, S.Y., Vester-christensen, M.B., Schjoldager, K.T.G., Kong, Y., Bennett, E.P., Mandel, U., Wandall, H., Levery, S.B., Clausen, H.. Mining the o-glycoproteome using zinc-finger nuclease – glycoengineered simple cell lines. Nat. Methods 2011;8:977-82.

2. Yang, Z., Steentoft, C., Hauge, C., Hansen, L., Thomsen,  a. L., Niola, F., Vester-Christensen, M.B., Frodin, M., Clausen, H., Wandall, H.H., Bennett, E.P.. Fast and sensitive detection of indels induced by precise gene targeting. Nucleic Acids Res. 2015;43:e59(1-8).

3. Duda, K., Lonowski, L. a, Kofoed-Nielsen, M., Ibarra, A., Delay, C.M., Kang, Q., Yang, Z., Pruett-Miller, S.M., Bennett, E.P., Wandall, H.H., Davis, G.D., Hansen, S.H., Frödin, M.. High-efficiency genome editing via 2A-coupled co-expression of fluorescent proteins and zinc finger nucleases or CRISPR/Cas9 nickase pairs. Nucleic Acids Res. 2014;42:e84(1-16).

4. Lonowski, L.A., Narimatsu, Y., Riaz, A., Delay, C.E., Yang, Z., Niola, F., Duda, K., Clausen, H., Wandall, H.H., Hansen, S.H., Bennett, E.P., Frödin, M. Genome editing using FACS enrichment of nuclease expressing cells and indel detection by amplicon analysis (IDAA). Nat. Protoc. 2016; in press.

5. Yeung, A.T., Hattangadi, D., Blakesley, L., Nicolas, E.. Enzymatic mutation detection technologies. Biotechniques 2005;38:749–758. 6. Overbeek, M. Van, Capurso, D., Carter, M.M., Thompson, M.S., Frias, E., Russ, C., Reece-hoyes, J.S., Nye, C., Gradia, S., Vidal, B., Zheng, J., Hoffman, G.R., Fuller, C.K., May, A.P.. DNA Repair Profiling Reveals Nonrandom Outcomes at Cas9-Mediated Breaks Article DNA Repair Profiling Reveals Nonrandom Outcomes at Cas9-Mediated Breaks. Mol. Cell 2016;63:1–14.