AID-seq In Vitro Off-Target Detection

AID-seq In Vitro Off-Target Detection

1. Background

From basic science to clinical application, gene-editing technology is redefining the boundaries of medicine at an unprecedented pace. In this revolutionary process, off-target effect assessment has become a critical bottleneck between scientific breakthroughs and clinical translation. Researchers require precise off-target data to reveal editing mechanisms and guide the development of next-generation high-fidelity tools; pharmaceutical R&D teams rely on comprehensive off-target analysis to evaluate candidate products and determine whether they can cross the clinical threshold; regulatory agencies regard off-target assessment as an insurmountable barrier to ensuring patient safety. These multidimensional needs are driving the rapid development of highly sensitive, unbiased off-target detection methods.

The FDA, in its guidance document Human Gene Therapy Products Incorporating Human Genome Editing (2024), explicitly requires comprehensive off-target risk assessment for gene-editing products, emphasizing the adoption of multiple methods for genome-wide analysis to reduce potential bias in off-target site identification. Among these, the FDA specifically highlights "biochemical" (biochemical-level detection) as exactly the experimental method based on in vitro DNA cleavage — an approach capable of fully detecting off-target sites of editing tools without the constraints of cellular environment or factors. Similarly, China's CDE, in the Technical Guidelines for Pharmacological Research and Evaluation of In Vivo Gene Therapy Products (Trial) (2022), aligns with the FDA in requiring multi-method identification of off-target events.

Figure 1. FDA and CDE guidance requirements for off-target detection

After years of continuous effort, ZhuHai GeneRulor has developed a novel high-throughput in vitro off-target detection method, AID-seq, which surpasses existing off-target detection methods in both sensitivity and specificity. More significantly, AID-seq introduces a high-throughput sgRNA screening strategy capable of simultaneously detecting the on-target/off-target profile of over a hundred sgRNAs in a single experiment, and builds a new CRISPR protein off-target prediction model based on large-scale data. To ensure technical reliability and accuracy, ZhuHai GeneRulor has conducted comprehensive method validation in strict accordance with ICH Q2(R1) guidelines and the FDA Guidance for Industry on Analytical Procedures and Methods Validation for Drugs and Biologics, establishing a complete technical evaluation framework. ZhuHai GeneRulor has successfully provided gene-editing safety evaluation services compliant with regulatory standards to numerous domestic and international gene therapy companies, fully supporting IND submissions and clinical translation.

2. Detection Principle

AID-seq is a highly efficient genome-wide off-target detection method. Its fundamental principle involves random fragmentation of genomic DNA and ligation with hairpin-structured i7 adapters to form a "dumbbell-shaped" structure. Subsequently, three different nuclease mixtures are used to digest the genomic DNA; this step is repeated 2–4 times to maximally remove false-positive signals from background DSBs. The "dumbbell-shaped" DNA with successfully ligated adapters on both ends is then incubated in a single-step reaction with the Cas9/Cas12a ribonucleoprotein (RNP) complex, exposing the freshly cleaved DNA termini, followed by ligation with biotinylated i5 adapters. The biotinylated adapter-ligated DNA is then enriched by streptavidin-coated magnetic beads. Finally, a multiplexed PCR approach is used to construct the final sequencing library.

Figure 2. Schematic illustration of the AID-seq off-target detection principle

3. AID-seq Detection and Analytical Advantages

3.1 Comprehensive Capture of DSB Sites Induced by Gene Editing

(1) In vitro cleavage: purified reagent components are used for in vitro experiments, improving detection reproducibility and avoiding the influence of cell transfection efficiency and cellular repair on detection; in vitro experiments can also substantially improve the detection sensitivity for low-frequency variant sites, facilitating the detection of low-frequency mutations;

(2) Genome-wide unbiased off-target detection: independent of bioinformatic predictions; background DSBs are eliminated through nuclease mixture processing, enabling the discovery of novel, previously unreported off-target sites;

(3) Low-frequency off-target event capture: high sensitivity achieved through multiple nuclease digestion and streptavidin enrichment, enabling detection of low-frequency off-target events and providing reliable data for sgRNA safety evaluation;

(4) Pooled AID-seq high-throughput screening: as a powerful tool for rapidly screening the optimal sgRNA, it enables simultaneous assessment of the off-target specificity of multiple sgRNAs, substantially improving screening efficiency and providing optimal guidance for CRISPR applications.

3.2 Outstanding Analytical Performance

3.2.1 Method Validation

ZhuHai GeneRulor rigorously follows international experimental standards, conducting systematic validation of AID-seq technology across four critical performance dimensions:

3.2.2 Comparison with Other Methods: Superior Sensitivity and Specificity

Through detection analysis of key on-target sites including VEGFA site 1, EMX1, and FANCF, it was found that: AID-seq ranks highest in the true off-target site detection rate validated by GUIDE-seq across all target sites (Figure 3B), with multiple sites achieving 100% detection — significantly superior to other methods. PR curves (Figure 3C) and ROC curves (Figure 3D) show that both the AUPRC and AUROC values of AID-seq are the highest, with performance significantly superior to the other three methods. This demonstrates that AID-seq can maximally reduce false positives while efficiently detecting true off-target sites, exhibiting far superior off-target detection sensitivity and specificity compared with existing methods, and this superiority remains stable across different target sites.

Figure 3. Comparison of off-target detection sensitivity and specificity: AID-seq vs. SITE-seq, CIRCLE-seq, and CHANGE-seq against GUIDE-seq validation

3.3 Service Advantages

(1) Comprehensive method validation: AID-seq has undergone systematic validation with rigorous quality control standards established at key workflow steps; the detection workflow strictly adheres to ICH Q2(R1) and FDA bioanalytical method validation guidelines, ensuring detection results comply with domestic and international regulatory standards;

(2) Leading technology platform: through the innovative "dumbbell-shaped" DNA structure combined with nuclease mixture processing, background DSBs from false positives are efficiently eliminated, ensuring the accuracy and reliability of detection results;

(3) Full-panel off-target site annotation: comprehensive multi-dimensional annotation information provided for off-target sites, including chromosomal positioning, gene regions, relationship to cancer genes, and more, enabling comprehensive assessment of off-target risk;

(4) Rich success case portfolio: AID-seq off-target detection services have been successfully delivered to numerous gene therapy companies, supporting smooth IND submissions;

(5) Customized analytical solutions: customized detection schemes based on target site characteristics and editing system (Cas9/Cas12 and others);

(6) Full-process service model: end-to-end gene-editing safety solutions encompassing target site design evaluation, sample testing, and IND submission data support.

4. Application Scenarios

(1) Target site screening-stage off-target risk assessment: genome-wide qualitative identification of off-target sites to help research teams screen for the target site with the lowest off-target risk;

(2) Basic research tool: an important tool for investigating the specificity mechanisms of the CRISPR-Cas system, driving the improvement and optimization of gene-editing technologies;

(3) Preclinical safety evaluation: meeting regulatory agency requirements for rigorous off-target detection of gene-editing therapeutics to provide key safety data for translational research;

(4) IND submission data support: provision of cell-based, genome-wide level off-target detection reports compliant with CDE and FDA requirements, supporting the clinical IND submission of gene-editing therapeutics.

In addition, Pooled AID-seq has the following application scenarios:

(1) Pooled AID-seq strategy can screen for sgRNAs with the highest cleavage efficiency and lowest off-target effects for gene editing and therapy;

(2) For CRISPR library optimization: CRISPR libraries are commonly used to identify genes related to drug sensitivity and resistance; however, libraries may contain sgRNAs with poor inclusivity, increasing unnecessary library size and causing strong mixing effects. Therefore, it is necessary to use pooled AID-seq strategy to screen for the best sgRNAs across the entire genome or a specific set of targets to reduce library size and improve library performance;

(3) AID-seq can be used as a general-purpose tool to characterize newly discovered CRISPR in a massively parallel manner with respect to cleavage efficiency, off-target effects, and PAM preference, and to build off-target prediction models based on these large-scale datasets.

5. Case Analysis

The AID-seq analytical reports provided by ZhuHai GeneRulor adopt a standardized format to ensure professional and comprehensive data presentation. The introductory section covers project background, experimental rationale, AID-seq technical principles, library construction workflow description, and bioinformatics analysis pipeline overview, providing clients with the necessary technical context. This is followed by sample information and sequencing data statistics, which include sample metadata, raw sequencing data quality metrics, and control group alignment result statistics, ensuring data quality and reliability. In addition, the report includes the following key components:

(1) Detailed on-target site information detected by AID-seq: recording the chromosome number, start and end genomic coordinates, strand orientation, and sequence information for each on-target site.

Figure 4. AID-seq representative report (on-target site detailed information)

(2) Off-target site basic information and functional annotation detected by AID-seq: the report provides detailed off-target analysis information from the gene-editing experiment, including sample name, chromosomal number, start coordinates, affected gene information and distance, gene region classification type (e.g., intergenic region, intronic), mismatch count, and whether cancer-associated genes are involved and other key data.

Figure 5. AID-seq representative report (off-target site detailed information)

(3) Off-target site visualization: the report provides the sgRNA sequence alignment status and off-target analysis for each gene-editing system in the sample. The top portion displays the reference sgRNA sequence; below it are the alignment results of each potential off-target site with that sequence. Each row represents one off-target site; color-coded markers indicate aligned bases; dots (.) represent bases identical to the sgRNA sequence; letters represent mismatched bases. The figure also includes the sequencing read count for each site. The visualization enables a direct presentation of the directional precision and off-target profile of the gene-editing tool, while simultaneously evaluating on-target efficiency and potential off-target risk.

Figure 6. AID-seq representative report (off-target site visualization)

6. Service Content

* Service turnaround: standard workflow 20–30 business days.

* Service highlight: integrated end-to-end service support from sgRNA design through final data analysis is available.

7. Sample Requirements

8. References

[1] Tian R, et al. Massively parallel CRISPR off-target detection enables rapid off-target prediction model building. Med, doi:10.1016/j.medj.2023.05.005 (2023).

[2] Cui Z, et al. FrCas9 is a CRISPR/Cas9 system with high editing efficiency and fidelity. Nat Commun 13, 1425, doi:10.1038/s41467-022-29089-8 (2022).