Ino-seq: ABE Off-Target Detection

Ino-seq: ABE Off-Target Detection

1. Background Introduction

As a crucial derivative technology of the CRISPR/Cas system, base editing (BE) enables precise base transitions (A·T→G·C or C·G→T·A) without inducing DNA double-strand breaks (DSBs), and has become a core tool in fields such as gene therapy and crop breeding [1,2]. Among these, adenine base editors (ABEs) can correct approximately 48% of known pathogenic point mutations, demonstrating enormous clinical application potential in the treatment of genetic diseases such as sickle cell anemia and β-thalassemia [2].

However, the off-target effects of ABEs remain a key bottleneck restricting their clinical translation—off-target events may lead to unintended consequences such as oncogene activation and tumor suppressor gene inactivation, seriously threatening the safety of gene therapy products [3,4]. The off-target effects of ABEs have dual origins: first, sgRNA-dependent off-targets, resulting from the tolerant recognition of similar genomic sequences by Cas proteins; second, sgRNA-independent off-targets, caused by the intrinsic activity of deaminase (TadA), which occur randomly at transcriptionally active regions or single-stranded DNA exposure sites [1,2].

With increasingly stringent regulatory requirements, the U.S. FDA clearly stipulates in the "Guidance for Industry: Gene Therapy Products" that gene therapy products must undergo systematic off-target effect assessment during preclinical research, clinical trials, and post-marketing monitoring [4]; China's NMPA-CDE also emphasizes the core status of off-target effect assessment in the "Technical Guidelines for Non-Clinical Research and Evaluation of Somatic Cell Therapy Products" [4]. However, existing methods for detecting off-target effects of adenine base editors still have three major limitations: first, methods relying on DSB intermediates or Cas9 binding events (including GUIDE-seq [5] and CHANGE-seq-BE [6]) can essentially only detect sgRNA-dependent sites, failing to capture the intrinsic, sgRNA-independent deamination events of adenine base editors; second, enrichment-based methods such as Selict-seq [2] and Tracking-seq [7] directly target editing products but exhibit unstable sensitivity and specificity, potentially missing low-frequency editing events or capturing non-specific signals; third, unbiased detection methods such as whole-genome sequencing (WGS) not only require extremely high sequencing depth (>100×) but also struggle to distinguish true editing events from background noise [8-10].

These limitations collectively highlight the necessity of developing a new method—one that can directly detect inosine with high sensitivity in a physiologically relevant cellular environment. To this end, we have developed Ino-seq, a genome-wide off-target site analysis method that can directly capture inosine-containing DNA generated by adenine base editors in living cells. By combining endonuclease V (EndoV)-mediated inosine cleavage with three complementary noise reduction strategies, Ino-seq can sensitively and specifically detect both sgRNA-dependent and sgRNA-independent off-target events.

2. Detection Principle

Ino-seq achieves genome-wide off-target detection by tracking deoxyinosine (dI) produced by ABEs (as shown in Figure 1). The core principle of this method is based on the specificity of endonuclease V (EndoV): EndoV recognizes inosine and precisely cleaves DNA at the second phosphodiester bond 3' to the inosine residue.

The specific workflow is as follows:

(1) Genomic DNA is extracted from ABE-treated cells and mechanically fragmented;

(2) Adapters are ligated to fragment ends, followed by damage repair and exonuclease treatment to remove DNA fragments without adapters;

(3) EndoV specifically introduces nicks at inosine-containing sites;

(4) After heat denaturation, single-stranded DNA binding protein (SSB) stabilizes the single-stranded structure. The cleaved single strands are not biotinylated, and uncleaved DNA fragments are removed using streptavidin magnetic beads to enrich inosine-containing DNA;

(5) Bridge adapters with unique molecular identifiers (UMIs) are ligated, followed by PCR amplification to complete library construction;

(6) During sequencing, the inosine site is consistently located at the second base downstream of the P7 sequencing primer.

Figure 1. Schematic Diagram of Ino-seq Detection Principle

3. Innovations and Advantages

3.1 Core Technological Innovations

(1) Direct inosine capture strategy: Directly targets inosine, the intermediate product of ABE editing, avoiding reliance on DNA double-strand breaks or Cas protein binding. This inherently ensures the specificity of off-target signals, with an enrichment efficiency of over 100-fold.

(2) Simultaneous coverage of dual off-target types: Can detect both sgRNA-dependent and sgRNA-independent off-target events, completely addressing the technical shortcoming of existing methods that can only capture a single type of off-target.

(3) UMI-mediated high-fidelity analysis: Original DNA fragments are labeled with unique molecular identifiers (UMIs) to effectively correct PCR amplification bias and sequencing errors. Combined with a control group background correction algorithm, the false positive rate is significantly reduced, and the validation rate of high-confidence sites exceeds 95%.

3.2 Key Performance Indicators

Compared with traditional methods, Ino-seq performs excellently in key indicators such as accuracy, precision, recall, F1 score, AUROC, and AUPRC, enabling highly specific capture of true off-target sites (as shown in Figure 2).

Figure 2. Performance Comparison of Four ABE Off-Target Detection Technologies

4. Application Scenarios and Service Advantages

4.1 Core Application Scenarios

(1) Preclinical safety evaluation of gene therapy products: Detects off-target effects of ABEs in cell lines, organoids, and animal models, providing off-target detection data compliant with FDA/NMPA regulatory requirements to support IND filing submissions [4].

(2) Optimization of base editor variants: Compares off-target profiles of different BE variants (e.g., ABE7.10, ABE8e, ABE8e (V106W)) to screen for high-specificity editors.

(3) Endogenous deaminase activity detection: Evaluates endogenous ADAR enzyme activity in different cell lines/tissues, providing a scientific basis for editing strategy optimization.

(4) Clinical trial sample monitoring: Tracks dynamic changes in off-target events in patient samples during clinical trials, comprehensively ensuring clinical application safety.

(5) Development and validation of research tools: Provides a standardized detection protocol for studying off-target characteristics of novel base editing technologies (e.g., high-fidelity ABEs).

4.2 Core Service Advantages

(1) Technical authority: Developed based on internationally recognized technical principles such as EndoV-seq, with performance indicators verified by internal core technical documents and invention patents, ensuring a rigorous and reliable technical route.

(2) Comprehensive compliance support: The reporting system strictly adheres to regulatory requirements such as the FDA's "Guidance for Industry: Gene Therapy Products" and NMPA's "Technical Guidelines for Non-Clinical Research and Evaluation of Somatic Cell Therapy Products," and can be directly used for IND filings and regulatory inspections.

(3) Standardized quality management: The laboratory operates an ISO9001 quality management system, and experimental processes strictly follow SOP specifications. All data are traceable and reproducible, meeting GLP experimental requirements.

(4) Professional technical team: Led by senior experts in the field of gene editing, core members have years of experience in off-target detection and data analysis, capable of providing customized experimental design and in-depth result interpretation.

(5) Efficient cycle guarantee: The standard service cycle is 20-30 working days, and the expedited service can be shortened to 15 working days to meet project schedule requirements; bilingual reports (Chinese and English) are provided to adapt to international filing scenarios.

5. Example Report

Shutong Technology provides standardized Ino-seq detection reports compliant with regulatory requirements, with core content including the following modules (data are examples; actual results shall prevail):

(1) On-Target Editing

Figure 3. Nucleotide Mutation at the On-Target Site

The nucleotide editing at the on-target site region is shown in Figure 3. The sgRNA binding position is marked at the top of the image; the upper part represents the control group (Blank), and the lower part represents the experimental group. The gray scale above the samples indicates the read count. Each read is represented by a gray block, with only mutated nucleotides displayed: light blue for A, red for G, dark blue for T, and orange for C.

(2) Off-Target Site Detection and Classification Statistics

Figure 4. Detection and Classification Statistics of Off-Target Sites

Figure 5. On-Target/Off-Target Site Mismatch Map

(3) Genomic Distribution and Functional Annotation of Off-Target Sites

The proportion of all significant off-target sites in each functional region of the genome is statistically analyzed and displayed as a pie chart (as shown in Figure 6).

Figure 6. Functional Region Annotation Statistics of Off-Target Sites

(4) Safety Risk Assessment

Off-target sites are annotated for oncogenes/tumor suppressor genes (referring to the COSMIC database), and editing frequencies are calculated, with an example shown in Figure 7:

Figure 7. Oncogene/Tumor Suppressor Gene Annotation of Off-Target Sites

6. Service Content

7. Sample Requirements

8. References

[1] Liang P, Xie X, Zhi S, et al. Genome-wide profiling of adenine base editor specificity by EndoV-seq[J]. Nature Communications, 2019, 10(1): 67.

[2] Yuan K, Xi X, Han S, et al. Selict-seq profiles genome-wide off-target effects in adenosine base editing[J]. Nucleic Acids Research, 2025, 53(7): gkaf281.

[3] Jin S, Zong Y, Gao Q, et al. Cytosine, but not adenine, base editors induce genome-wide off-target mutations in rice[J]. Science, 2019, 364(6437): eaaw7166.

[4]国家药品监督管理局药品审评中心。体细胞治疗产品非临床研究与评价技术指导原则[Z].2017.

[5] Tsai S Q, Zheng Z, Nguyen N T, Joung J K, et al. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases[J]. Nature Biotechnology, 2015, 33(2): 187-197.

[6] Lazzarotto C R, Miller S M, Kleinstiver B P. Sensitive and unbiased genome-wide profiling of base-editor induced off-target activity using CHANGE-seq-BE [J]. Nature Biotechnology, 2023, 41 (11): 1646-1653.

[7] Zhu M, Wienert B, Schaefer M, et al. Tracking-seq reveals the heterogeneity of off-target effects in CRISPR–Cas9-mediated genome editing[J]. Nature Biotechnology, 2024, 43(7): 799-810.

[8] Nahmad A D, Oren Y, Sorek R. Frequent aneuploidy in primary human T cells after CRISPR–Cas9 cleavage[J]. Nature Biotechnology, 2022, 40(12): 1807-1813.

[9] Leibowitz M L, Lee S, Kweon J, et al. Chromothripsis as an on-target consequence of CRISPR-Cas9 genome editing[J]. Nature Genetics, 2021, 53(10): 895-905.

[10] Kosicki M, Tomberg K, Bradley A. Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements[J]. Nature Biotechnology, 2018, 36(8): 765-771.