In-vivo detection method for synchronously detecting CRISPR (clustered regularly interspaced short palindromic repeats) off-target sites and viral vector integration sites

Abstract

The invention relates to the technical field of crossing of molecular biology, nucleic acid detection and bioinformatics, and discloses an in-vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites. According to the in-vivo detection method for synchronously detecting the off-target site of the CRISPR and the integration site of the virus vector, the reverse terminal repeat (ITR) of the AAV vector is used as a tag sequence for the first time, and a universal PCR (Polymerase Chain Reaction) amplification primer is designed and a unique molecular identifier (UMI) technology is introduced to detect the off-target site of the CRISPR and the integration site of the virus vector. And meanwhile, an off-target event and an AAV carrier integration event in an in-vivo CRISPR editing process are captured and accurately recognized. Compared with the existing schemes such as GUIDE-seq, DISCOVER-Seq +, OliTag-seq, U-DiTaS and the like, the in-vivo detection method for synchronously detecting the CRISPR off-target site and the virus vector integration site, provided by the invention, has the advantage that remarkable breakthrough is realized in the aspects of in-vivo applicability, detection depth, operation economy and industrialization standardization.

Description

Technical Field

[0001] The present invention relates to the interdisciplinary field of molecular biology, nucleic acid detection and bioinformatics, and in particular to an in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites. Background Art

[0002] Currently, safety evaluation of in vivo gene editing and adeno-associated virus (AAV) vector-based gene therapy focuses on two aspects:

[0003] 1) Distribution of vector genome integration sites across the genome;

[0004] 2) Potential off-target effects caused by double-strand DNA breaks (DSBs) induced by tools represented by the CRISPR / Cas series of nucleases.

[0005] To achieve the above goals, various detection methods based on high-throughput sequencing have been proposed in the prior art. However, these methods still have certain limitations in terms of labeling strategy, detection range, and sensitivity, as follows:

[0006] ITR-seq method: This method uses AAV terminal repeats (ITRs) as single-side primers to enrich and locate integration sites through short-read sequencing. Its characteristics include: enrichment of ITR-host chimeric sequences without the introduction of exogenous oligonucleotide tags; reliance on the paired-end Illumina sequencing platform, with a typical read length of 150-250bp; the lack of unique molecular identifiers (UMIs) makes it susceptible to PCR amplification bias in low-frequency events (<1%); and its primary application is the analysis of AAV vector integration events, not simultaneously capturing off-target sites caused by DSBs.

[0007] GUIDE-tag method: This method is based on the GUIDE-seq principle. It integrates double-stranded oligonucleotides (dsODN) at DSB sites, combining Tn5 enzyme-mediated library construction with UMI technology to achieve off-target detection. Its characteristics include: off-target detection sensitivity is approximately 3-5 times higher than the original GUIDE-seq method, and it can be used for off-target detection in in vivo tissues; it requires additional delivery of dsODN to cells, which is relatively complex and may interfere with editing efficiency; it cannot detect AAV vector integration events; and due to the short sequencing read length, it is difficult to effectively resolve variants such as complex insertions, inversions, or large deletions.

[0008] UDiTaS method: This method combines site-specific primer PCR with Tn5 adapter library construction, enabling simultaneous assessment of multiple variants, including insertions and deletions (Indels), inversions, and chromosomal translocations, at the same locus. Features of this technology include: simultaneous detection of multiple types of structural variation using a single library workflow; however, its reliance on site-specific primers prevents unbiased detection across the entire genome; the lack of ITR-specific tag design makes it unsuitable for AAV integration site monitoring; and the lack of UMIs during library construction makes quantitative results susceptible to PCR amplification cycling.

[0009] Other related methods: New technologies including OliTag-seq and AID-seq have improved the sensitivity of off-target detection, but they generally rely on exogenous tags or in vitro enzyme cleavage, and it is difficult to simultaneously monitor viral vector integration and genomic off-target events.

[0010] From the above, it can be seen that the existing technology still lacks a comprehensive detection method that does not require the introduction of exogenous tags, can simultaneously obtain AAV integration information and DSB off-target data in a single sequencing, and has the ability to debias UMI.

[0011] To address the above shortcomings, the present invention fully utilizes the natural ITR sequence of the vector as an enrichment primer, embeds an 8-12nt UMI tag upstream of the adapter to eliminate PCR amplification bias, and uses high-throughput short-read sequencing to perform detection in a single sequencing process of a single library. It can detect low-frequency events with a frequency of ≥0.1% with a sensitivity of ≥95% within <48 hours, taking into account both quantitative accuracy and structural information analysis, providing a solution for the comprehensive safety assessment of therapeutic gene editing products. Summary of the Invention

[0012] The purpose of this invention is to provide an in vivo detection method that complies with the latest regulatory requirements such as FDA / ICH-M4 and can analyze the dual risks of "editing off-target + vector integration" in parallel within 48 hours.

[0013] To achieve the above object, the present invention provides the following technical solutions:

[0014] One of the technical solutions of the present invention:

[0015] An in vivo method for simultaneously detecting CRISPR off-target sites and viral vector integration sites comprises the following steps:

[0016] S1. Construct an AAV vector carrying an inverted terminal repeat sequence tag and co-deliver it with the CRISPR editing element to the cells or animals to be tested;

[0017] S2. Collect cells or animal tissues, extract genomic DNA and fragment it;

[0018] S3. After ligating the fragmented DNA to sequencing adapters, two rounds of PCR amplification were performed using first-round PCR primers containing unique molecular identifiers and second-round PCR primers containing universal ITR primers to obtain a sequencing library;

[0019] S4. Use a paired-end high-throughput sequencing platform to obtain sequencing data;

[0020] S5.UMI clustering removes repetitive amplified sequences, and then the Smith-Waterman local sequence alignment algorithm is used to accurately realign the structural fragments, precisely identifying off-target sites and AAV viral vector integration sites.

[0021] Furthermore, in step S1, the serotype of the AAV vector is one of AAV2, AAV6, AAV8, AAV9 or an engineered variant thereof, and the genome length of the AAV vector is ≤4.0 kb.

[0022] Furthermore, in step S1, the CRISPR editing element is one of SpCas9, eSpCas9, SaCas9, AsCas12a, LbCas12a, Prime Editor or Base Editor.

[0023] Furthermore, in step S2, the fragmentation is to fragment into 300 to 800 bp.

[0024] Furthermore, in step S4, the use of a double-end high-throughput sequencing platform to obtain sequencing data specifically includes: using the Illumina platform for 150bp double-end sequencing and the Nanopore platform LSK114 for long-read sequencing.

[0025] Furthermore, in step S5, the UMI clustering to remove duplicate amplified sequences is specifically as follows:

[0026] a. Data preprocessing

[0027] The raw FASTQ files obtained from sequencing data were quality controlled to remove low-quality sequences and sequencing adapter sequences;

[0028] b.UMI extraction and preliminary clustering:

[0029] In sequencing data, each sequence starts with a UMI sequence of 8 to 12 bp in length, which is used to label the original template molecule for PCR amplification. First, the UMI tag of each sequence is extracted, and preliminary clustering and grouping are performed based on the UMI tag.

[0030] c.UMI clustering deduplication algorithm:

[0031] Based on the characteristics of the Illumina platform, a UMI clustering algorithm based on the edit distance threshold is used to cluster sequence UMIs. The threshold is optimized to an edit distance of 1, which means that only UMIs with a single base difference are allowed to be merged into the same cluster unit to effectively eliminate amplification bias caused by the PCR process. Each UMI cluster unit retains a representative sequence.

[0032] d. Structural decoding analysis and double-end sequence splicing:

[0033] The double-ended sequences after de-redundancy treatment were spliced ​​using Flash or PEAR tools. The overlapping region length after splicing was required to be ≥10bp and the mismatch rate was ≤0.1% to ensure the accuracy of the spliced ​​sequence. The complete sequence after splicing was used for the next analysis.

[0034] Furthermore, in step S5, the Smith-Waterman local sequence alignment algorithm is used to accurately realign the structural fragments as follows:

[0035] The sequences were aligned with the reference genome, and the optimization parameters were set as follows: match score +2, mismatch penalty -2, gap opening penalty -3, and gap extension penalty -1. After sequence alignment, the sequences were analyzed to determine whether they contained the characteristic structure of the AAV-ITR junction with the genome.

[0036] Furthermore, in step S5, the precise identification of off-target sites and AAV viral vector integration sites is specifically:

[0037] A dual-index strategy was designed, with one end anchoring the AAV-ITR sequence features and the other end anchoring the genomic sequence features, to achieve accurate distinction and identification of off-target sites and AAV integration sites.

[0038] The second technical solution of the present invention:

[0039] An in vivo detection system for simultaneously detecting CRISPR off-target sites and viral vector integration sites based on an in vivo detection method for simultaneously detecting CRISPR off-target sites and viral vector integration sites, comprising:

[0040] A data preprocessing module that supports direct import of paired-end FASTQ files from the Illumina sequencing platform and automatically performs data quality control using built-in quality control tools, including adapter sequence removal and low-quality sequence filtering.

[0041] Demultiplex and UMI deduplication module, which automatically separates and identifies samples by reading index sequence files and accurately extracts UMI sequences. It also uses an efficient UMI clustering algorithm to remove redundant PCR amplification sequences to ensure high data accuracy.

[0042] Sequence alignment and off-target detection module, which uses the Smith-Waterman local alignment algorithm to achieve accurate alignment and quickly identify and mark potential off-target sites. This module automatically calculates the AviTag-seq score for each site and generates a visual chart to intuitively display the distribution of off-target events.

[0043] AAV integration site analysis module, which uses an efficient algorithm to accurately identify the genomic integration site of the AAV vector and generates a detailed integration site analysis report by annotating the integration site;

[0044] An automated report generation module that automatically integrates analysis results into standardized PDF or HTML reports.

[0045] Furthermore, the report content includes: sample quality control information, a detailed list of off-target sites and integration sites, visual charts, experimental parameters and analysis process information.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention provides an in vivo detection method for simultaneously detecting CRISPR off-target sites and viral vector integration sites. For the first time, the inverted terminal repeat sequence (ITR) of the AAV vector itself is used as a tag sequence. By designing universal PCR amplification primers and introducing unique molecular identifier (UMI) technology, it can simultaneously capture and accurately identify off-target events and AAV vector integration events during the CRISPR editing process in vivo.

[0048] The in vivo detection method for simultaneous detection of CRISPR off-target sites and viral vector integration sites provided by the present invention has achieved significant breakthroughs in in vivo applicability, detection depth, operational economy, and industrial standardization compared to existing methods such as GUIDE-seq, DISCOVER-Seq+, OliTag-seq, and U-DiTaS. Specifically, it is embodied in the following aspects:

[0049] High in vivo multi-tissue stability: The method has been validated in multiple in vivo tissues, including mouse liver, spleen, bone marrow, and brain, with an ITR tag-positive detection rate of ≥80%. The sequencing read consistency between different tissues was R2 ≥0.92, and the coefficient of variation (RSD) was ≤8%.

[0050] Simultaneously output dual risk data in a single experiment: Leveraging ITR single-anchor primers and UMI debiasing strategies, a single library and single sequencing run are required to simultaneously analyze CRISPR off-target DSBs and AAV / IDLV integration sites. Results are automatically generated in regulatory-friendly formats, including a Circos plot and a CSV / VCF table, meeting FDA e-CTD / NMPAIND attachment requirements.

[0051] Simplified process and universal compatibility: ITR universal primers are compatible with over 95% of commercial AAV serotypes on the market, eliminating the need for strand-specific primer design. The method is compatible with a variety of gene editing tools (such as SpCas9, SaCas9, As / LbCas12a, Cas12f, PrimeEditor, BaseEditor, etc.). It is applicable to different delivery methods such as VLP, IDLV, and LNP-mRNA, requiring only a change of vector to maintain a unified amplification and detection process.

[0052] Significantly improved sensitivity and accuracy: In the AsCas12a system, the method of the present invention increases the number of valid reads by 15 to 20 times compared with DISCOVER-Seq+ / OliTag-seq; the minimum detectable event frequency reaches 0.1%; the comprehensive recognition F1-score is ≥0.90, with a sensitivity of ≥0.88 and a specificity of ≥0.93, which is significantly better than existing in vivo detection technologies;

[0053] Low-cost, high-efficiency, and standardized output: End-to-end time from sample processing to result reporting is ≤ 48 hours (including DNA extraction, library preparation, 250-cycle PE150 sequencing, and automated analysis); the total cost of a single sample test is ≤ US$200 (including sequencing and consumables), saving ≥ 70% compared to DISCOVER-Seq+; The system comes with standard operating procedures (SOPs), a cGMP documentation package, and a traceable log. Both graphical user interface (GUI) and command line interface (CLI) software are available, allowing for one-click generation of PDF / HTML / JSON reports, facilitating integration into LIMS and QA systems.

[0054] Industrialization and Regulatory Prospects: The analytical platform has passed ISO13485 software lifecycle documentation audits and can be seamlessly integrated into CDMO / CRO pipelines. It supports batch analysis and API calls, meeting the parallel testing needs of high-throughput drug pipelines. It provides comprehensive safety data on "off-target and vector integration" for novel gene therapy products, significantly reducing the technical barriers to IND filing and broadening commercial application opportunities.

[0055] In summary, the in-vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites provided by the application has the comprehensive advantages of high sensitivity, double-index synchronous detection, low cost, and process standardization, and lays a foundation for its core competitiveness in the field of safety evaluation of gene therapy and cell engineering.

[0056] The in-vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites provided by the application fills the blank in the field of synchronous detection of in-vivo off-target effects and viral vector integration events, and provides a new efficient and reliable method for safety evaluation of gene therapy technology. BRIEF DESCRIPTION OF DRAWINGS

[0057] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are intended to only illustrate preferred embodiments and are not intended to limit the application. Moreover, like reference numerals are intended to denote like parts throughout the various drawings. In the drawings:

[0058] Figure 1 It is a schematic diagram of AAV viral vector integration mechanism;

[0059] Figure 2 It is a schematic diagram of ITR universal primer amplification principle;

[0060] Figure 3 It is a schematic diagram of two-round PCR amplification process;

[0061] Figure 4 It is a UMI structure decoding algorithm flowchart;

[0062] Figure 5 It is a schematic diagram of the overall process of the method provided by the application;

[0063] Figure 6 It is a detection success rate heat map of the application in various tissues of mice;

[0064] Figure 7 It is a comparison chart of biological repeated experiment results of three independent experimental batches;

[0065] Figure 8 It is a Wayne diagram comparison of off-target detection results of three experimental batches;

[0066] Figure 9 It is a data analysis result chart of repeated experiments of the same sample;

[0067] Figure 10 It is a correlation analysis result chart between off-target event sequencing read numbers and gene editing efficiency;

[0068] Figure 11Visualization of the chromosomal distribution of off-target and viral integration events. DETAILED DESCRIPTION

[0069] Various exemplary embodiments of the present invention are now described in detail. This detailed description should not be considered as a limitation of the present invention, but should be understood as a more detailed description of certain aspects, features, and embodiments of the present invention. It should be understood that the terms used in the present invention are only for describing specific embodiments and are not intended to limit the present invention.

[0070] In addition, for numerical ranges in the present invention, it is understood that each intervening value between the upper and lower limits of the range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated value or intervening value in the stated range is also included in the present invention. The upper and lower limits of these smaller ranges may independently be included or excluded in the range.

[0071] Unless otherwise indicated, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art. Although only preferred methods and materials are described herein, any methods and materials similar or equivalent to those described herein may also be used in the practice or testing of the present invention. All documents mentioned in this specification are incorporated by reference to disclose and describe the methods and / or materials associated with the documents. In the event of any conflict with any incorporated document, the contents of this specification shall prevail.

[0072] It will be apparent to those skilled in the art that various modifications and variations may be made to the specific embodiments described herein without departing from the scope or spirit of the invention. Other embodiments will be apparent to those skilled in the art from the description of the invention. The description and examples are intended to be exemplary only.

[0073] The words “include,” “including,” “have,” “contain,” etc. used in this document are open-ended terms, meaning including but not limited to.

[0074] Example 1

[0075] Construction of AAV ITR-tagged vector

[0076] 1) Plasmid design and construction

[0077] Using AAV2 serotype as the vector backbone, retaining the 145bp ITR sequence, a 39bp random tag sequence (Tag-39) was precisely inserted 60bp upstream of the right ITR on the plasmid pAAV-CMV-EGFP backbone using Gibson Assembly technology to ensure that the tag read can be uniquely located in the final sequencing;

[0078] 2) Virus packaging and preparation

[0079] HEK293T cells (ATCC No.: CRL-3216) were used for adeno-associated virus packaging, which specifically included the following steps:

[0080] □ Using PEI MAX as the transfection reagent, co-transfect HEK293T cells with the pAAV vector plasmid containing the ITR tag sequence, the helper plasmid pHelper, and the packaging plasmid pRC6 using the three-plasmid co-transfection method. After transfection, culture the cells at 37°C, 5% CO2 for 72 hours.

[0081] 3) Virus purification and quantification

[0082] Viral particles were purified from the harvested cells using 15% to 60% iodixanol density gradient ultracentrifugation, and the virus was further concentrated by tangential flow filtration (TFF). The virus titer was then determined using real-time fluorescence quantitative PC, and the final viral vector titer was 1.3 × 10 13 vg / mL; endotoxin test results <1EU / mL, which meets the standards for subsequent animal experiments.

[0083] Functional verification of the AAV ITR tag vector prepared in Example 1:

[0084] Preliminary experiments were conducted in the Cas9 expression system of HEK293T cells. Through fluorescence microscopy observation and PCR detection, the results confirmed that the integration positive rate of the ITR tag sequence reached 86±4% (n=3, independent repeated experiments), demonstrating stable and efficient vector genome integration ability, providing reliable data support for subsequent in vitro and in vivo experiments.

[0085] Example 2:

[0086] Combined detection of off-target and integration events in vitro (HEK293T and K562 cells)

[0087] 1) Co-delivery of CRISPR and AAV vectors

[0088] The CRISPR editing element (SpCas9 or AsCas12a protein, 1 μg) was combined with the AAV ITR tag vector (2×10 4 vg / cell) were mixed and then The transfection system is delivered to target cells through electroporation;

[0089] Among them, HEK293T cells used procedure A-023, and K562 cells used procedure T-016;

[0090] 2) Genomic DNA extraction and fragmentation

[0091] Cells were collected 72 h after transfection, and genomic DNA was extracted using the Qiagen Gentra Puregene Blood Kit. The genomic DNA was then randomly fragmented to an average fragment length of 500–700 bp using a Covaris S220 sonicator (power 110 W, duty cycle 5%, cycle number 200, duration 80 s).

[0092] 3) Two-round PCR amplification and UMI labeling

[0093] Adapter ligation: The fragmented DNA is end-repaired and A-tailed, and then connected to the pre-designed sequencing adapter oligonucleotides;

[0094] The adapter sequence contains the Illumina flow cell binding sequence and a random UMI index region;

[0095] First-round PCR amplification: A pair of primers containing the UMI region and an ITR primer (ITR-R1) are used to perform the first round of amplification on the adapter-ligated DNA product. This introduces the UMI tag at the molecular level to achieve debiasing and simultaneously enrich the AAV-ITR chimeric sequence contained in the amplified fragment.

[0096] The first-round PCR reaction system is shown in Table 1;

[0097] Table 1 PCR reaction system

[0098]

[0099] The first round of PCR amplification procedure uses a gradual cooling strategy:

[0100] Initial denaturation at 98°C for 5 seconds; annealing starting at 68°C for 15 seconds, decreasing by 1°C each cycle for a total of 8 cycles. After the annealing temperature dropped to 60°C, annealing was continued at 60°C for 15 seconds; extension at 72°C for 30 seconds for a total of 22 cycles;

[0101] The length of the first-round amplification product is approximately 0.3 kb;

[0102] Second round of PCR amplification

[0103] A second round of amplification using a barcoded index primer and an ITR primer (ITR-R2) utilizes the reverse complementary nature of the AAV ITR sequence ends to simultaneously amplify all integration event fragments on both chains, avoiding the complexity of chain-specific primer design in traditional methods.

[0104] The barcoded primers used in the second round of PCR are shown in Table 2 ;

[0105] Table 2 barcoded primers used in the second round of PCR (5'→3')

[0106]

[0107]

[0108] 4) Sequencing library construction and sequencing

[0109] After purifying the second round of PCR amplification products, library construction was performed, and Illumina platform was used for 150bp double-end sequencing (PE150) to ensure sufficient depth and accurate and reliable data;

[0110] The integration of AAV viral vectors and the principle of amplification of ITR single-sided universal primers are shown in Figure 1 ;

[0111] Figure 1 The precise integration mechanism of AAV viral vectors near the CRISPR-induced DNA double-strand break site and the binding region of universal AAV ITR primers in the amplification process are clearly shown; by using the natural structural characteristics of ITR sequences to design single-sided primers, the present application successfully avoids the problem that traditional methods are prone to failure in high GC content regions;

[0112] The schematic diagrams of two rounds of PCR amplification are shown in Figure 2 and Figure 3 , wherein Figure 2 schematically shows the enrichment principle of ITR single-sided primers for vector integration fragments in the host genome, Figure 3 schematically shows the process of sample library construction through two rounds of PCR amplification and the introduction of exponential barcodes;

[0113] 5) Data analysis and algorithm decoding

[0114] After obtaining the sequencing data, first, UMI clustering is performed to remove duplicate amplified sequences, then Smith-Waterman local sequence alignment algorithm is used for precise re-alignment analysis of the detected structural variation fragments, and finally a standard variation format (VCF) result file is output, which accurately identifies potential off-target sites and AAV viral vector integration sites;

[0115] UMI clustering and structure decoding algorithm:

[0116] The data analysis process of the present application optimizes the unique molecular identifier (UMI) clustering and structure decoding analysis method for Illumina platform 150bp double-end sequencing data, and the specific process is as follows:

[0117] Data preprocessing: The original FASTQ files obtained by sequencing were imported into the analysis software for automatic quality control, and low-quality sequences (Phred quality value Q < 30) and sequencing adapter sequences were removed;

[0118] UMI extraction and preliminary clustering: Each sequencing read contains an 8-12bp UMI tag at the beginning of the sequence, which is used to mark the original template molecule from which it originated. The analysis software first extracts the UMI sequence of each sequence and performs preliminary clustering grouping based on the UMI.

[0119] UMI clustering and deduplication: A UMI clustering algorithm based on an edit distance threshold is used to perform fine-grained clustering and deduplication on the initially grouped sequences. To address the characteristics of the second-generation sequencing (Illumina) platform, the UMI edit distance threshold is set to 1, meaning that only UMIs with a maximum base difference are allowed to be clustered together, effectively eliminating bias introduced by the PCR amplification process. Each UMI cluster retains only one representative sequence (usually the most frequently occurring sequence in the cluster), thereby reducing the amount of duplicate data while maximizing the preservation of true variant signals.

[0120] Structural decoding analysis and read assembly: After deduplication, paired-end reads were assembled into a single sequence using Flash or PEAR. The overlap region of the assembled sequence was required to be no less than 8–15 bp, and the mismatch rate in the overlap region was required to be no more than 0.1%. These parameters ensured the accuracy of the assembly results when the reads were short. If some reads were long and lacked overlap, the unassembled paired-end sequences were retained for subsequent analysis.

[0121] Local alignment and structural variation decoding: The Smith-Waterman algorithm was used to perform local sequence alignment analysis on the spliced ​​sequences and align them with the reference genome sequence. The initial optimization settings for the alignment parameters were: match score +2, mismatch penalty -2, gap opening penalty -3, and gap extension penalty -1. Further testing showed that adjusting the parameters within the range of match score +1 to +3, mismatch penalty -1 to -3, gap opening penalty -2 to -4, and gap extension penalty -1 to -2 had little impact on the accuracy of the alignment results, indicating that the algorithm is robust to parameter changes. After sequence alignment, the sequences were analyzed to determine whether the characteristic structure of the AAV-ITR junction with the host genome existed.

[0122] Dual-index strategy for identifying variant types: Utilizing a pre-designed dual-index (dual-anchor) identification strategy, one end indexes the AAV ITR tag sequence, while the other end indexes the reference genome sequence. By detecting the positional relationship between the anchor sequences at both ends of the sequence, it is possible to accurately distinguish between AAV-host integration event fragments (sequences containing both ITR residues and genomic sequence) and pure genomic DSB repair fragments (sequences containing only genomic information without ITR residues), thereby achieving simultaneous identification of vector integration sites and off-target sites.

[0123] Verification of algorithm accuracy and sensitivity: The performance of the aforementioned algorithm pipeline was evaluated using both synthetic standard sequences and simulated datasets. The results demonstrated that the accuracy of the UMI clustering deduplication algorithm exceeded 99%, and the structural decoding analysis achieved over 95% consistency between simulated and real experimental data, effectively identifying low-frequency events with frequencies as low as 0.1%.

[0124] Software Architecture and Implementation: The data analysis process software of the present invention adopts a modular design, including functional modules such as data preprocessing, UMI clustering and deduplication, sequence assembly, local sequence alignment and decoding, off-target and integration site identification, and result visualization. It is implemented in Python and Shell scripts. The software has a user-friendly interface, supports the Linux operating system, and has efficient computing performance and easy operation.

[0125] The flowchart of the UMI structure decoding algorithm is as follows Figure 4 shown.

[0126] In order to further improve the applicability and flexibility of the method of the present invention, the key technical parameters in the data analysis process were expanded and optimized;

[0127] UMI length parameter expansion: The present invention initially uses a UMI length of 8-12 bp. By evaluating the impact of UMI lengths of 6 bp, 8 bp, 10 bp, 12 bp, and 14 bp on sequencing deduplication, the results show that UMI lengths of 8-12 bp can effectively distinguish a sufficient number of original template molecules, with a deduplication efficiency exceeding 99%. When the UMI length is less than 8 bp, the probability of misclassification between different template molecules increases significantly (the misclassification rate increases by approximately 5%). However, when the UMI length is extended to more than 12 bp, the deduplication effect does not improve significantly and the sequencing cost increases. Therefore, the present invention prefers to use UMIs of 8-12 bp, but the UMI length can be flexibly adjusted within the range of 6-14 bp to meet specific experimental requirements.

[0128] Extension of overlap parameters: The overlap length for double-end sequence splicing was originally set to ≥10 bp and a mismatch rate ≤0.1%. Experimental evaluation found that overlap lengths between 8 and 15 bp maintained high splicing accuracy. When the overlap length was less than 8 bp, the splicing sequence error rate increased significantly, while splicing efficiency did not improve further after exceeding 15 bp. Therefore, it is preferred to set the overlap length range to 8-15 bp and the mismatch rate to less than 0.1% to balance splicing success rate and accuracy.

[0129] Local alignment algorithm parameter expansion: This paper uses the Smith-Waterman algorithm for local sequence alignment, with the original optimized parameters of +2 for matches, -2 for mismatches, -3 for gap openings, and -1 for gap extensions. Testing within a wider parameter range showed that when the match score ranged from +1 to +3, the mismatch penalty ranged from -1 to -3, the gap opening penalty ranged from -2 to -4, and the gap extension penalty ranged from -1 to -2, this method was able to maintain high alignment accuracy without significant performance degradation. This demonstrates that the local alignment algorithm is robust to certain parameters and can be adjusted appropriately according to different experimental requirements.

[0130] This example successfully detected both gene editing off-target sites and random AAV vector integration sites at the cellular level through the aforementioned process, and quantitatively analyzed the frequency of each event. The analysis results were generated into a standard VCF format file and a graphical report, verifying the effectiveness and accuracy of the method.

[0131] Example 3

[0132] Combined detection of off-target and viral integration in mice

[0133] 1) Animal model and treatment

[0134] Six- to eight-week-old C57BL / 6J mice (half male and half female) were used as experimental animals and housed in a specific pathogen-free (SPF) animal facility. Four hours before gene-editing vector administration, mice were pre-immunosuppressed with intraperitoneal injections of cyclophosphamide and prednisone (50 mg / kg each) to minimize interference of the host immune response with subsequent testing.

[0135] 2) Vector and gene editing system delivery

[0136] Gene editing tools and AAV vectors were delivered in vivo by rapid injection into the mouse tail vein (completed within approximately 5 seconds);

[0137] Specific plan: AAV8 serotype vector is used to package SpCas9 expression vector, and the dosage is 1×10 12 vg / kg; AAV9 serotype vector is used to package sgRNA expression vector, and the dosage is 3×1012 vg / kg; the two vectors were mixed and injected into mice once;

[0138] 3) Sample collection and pretreatment

[0139] Three weeks after vector injection, mouse tissue samples, including liver, spleen, and bone marrow (approximately 50 mg each), were collected. Genomic DNA from each tissue was extracted using the Qiagen Gentra Puregene Blood Kit. Subsequent steps, including DNA fragmentation, adapter ligation, two-round PCR amplification, and library construction, were the same as in Example 2 (except that different barcoded primers were used for different tissue samples in the second round of PCR). After obtaining sequencing libraries, all sample libraries were subjected to high-throughput PE150 sequencing on the Illumina platform, and the data were entered into analysis software for automatic analysis according to the method and process of Example 2.

[0140] The overall flow diagram of the method provided by the present invention is as follows Figure 5 As shown;

[0141] Figure 5 The complete technical process of the AviTag-seq method was demonstrated in detail, including the co-delivery of AAV viral vectors and CRISPR systems, insertion of ITR tag sequences, sample tissue collection, genomic DNA extraction, amplification of ITR-flanked universal primers combined with UMI labeling, high-throughput sequencing (Illumina platform), and data structure analysis and automated report generation, demonstrating the efficiency and simplicity of this method.

[0142] The statistical results of the positive detection rates of ITR tag sequences in each tissue sample are shown in Table 3 ;

[0143] Table 3 Tag positive detection rate

[0144]

[0145] As shown in Table 4, the method provided by the present invention has high consistency and sensitivity in the detection results of different animal tissues; statistical analysis shows that the consistency of the results between different tissues is R 2 =0.92, with a coefficient of variation (RSD) of ≤8% for each tissue assay. Compared with existing methods (e.g., the improved DISCOVER-Seq+, which typically has a positive detection rate of less than 60% in vivo), the method of the present invention has significantly higher sensitivity for detecting the same sgRNA target. This indicates that the method of the present invention has significant advantages in sensitivity and accuracy in an in vivo setting.

[0146] Figure 6 A heat map of the detection success rate of the method of the present invention in various mouse tissues is provided;

[0147] Depend onFigure 6 It can be seen that the positive detection rate of this method in different tissues (such as liver, spleen, bone marrow, etc.) remains at a high level, highlighting the consistency of results and wide applicability of the AviTag-seq method in in vivo applications, and further proving the reliability of this method;

[0148] To further evaluate the robustness and reproducibility of the method of the present invention, we conducted validation under different experimental batches, different operators, and different laboratory conditions;

[0149] Robustness of different experimental batches: For the target gene site EMX1, three experimental batches of biological replicates were independently performed; the test results of the three batches are as follows Figure 7 As shown;

[0150] Depend on Figure 7 It can be seen that in three independent batches, the AviTag-seq method can stably identify high-abundance off-target sites. The sequencing reads of corresponding off-target sites are highly consistent between batches, demonstrating the reliable sensitivity of the AviTag-seq method.

[0151] Consistency of biological replicates: The off-target site identification results of different batches of biological replicates were analyzed by Venn diagrams. The Venn diagrams of the three batches were drawn as follows: Figure 8 As shown;

[0152] Depend on Figure 8 A total of 26 off-target sites were consistently detected across all replicates across the three batches. This high consistency strongly demonstrates the high reproducibility and reliability of the AviTag-seq method in biological replicates.

[0153] Stability of repeated experiments: To further confirm the stability of the AviTag-seq method, the same sample was tested twice independently. The data analysis results of the repeated experiments are shown in Figure 2. Figure 9 As shown;

[0154] Depend on Figure 9 It can be seen that the off-target site frequencies measured between different repeated experiments are highly correlated (Pearson correlation coefficient R 2 =0.99, P <0.0001), further demonstrating the excellent robustness and reliability of AviTag-seq technology;

[0155] Quantification capability: Further analysis of the correlation between the number of sequencing reads of off-target events detected by AviTag-seq and the frequency of insertions / deletions (Indels) caused by CRISPR-Cas9 editing; Figure 10 Shown are the results of the correlation analysis between off-target reads and gene editing efficiency;

[0156] Depend onFigure 10 It can be seen that the number of off-target reads detected by AviTag-seq is significantly positively correlated with the gene editing efficiency (R 2 =0.41, P=0.04); This shows that the method of the present invention can not only accurately detect the occurrence of off-target events, but also indirectly reflect the efficiency of gene editing through readout, and has a certain quantitative evaluation capability;

[0157] Distribution of off-target and integration events: Visualization of the distribution of off-target and AAV virus integration events on chromosomes. Figure 11 (Circos diagram) shown;

[0158] Depend on Figure 11 It can be seen that the Circos diagram clearly and intuitively shows the location distribution of CRISPR off-target events and AAV virus integration events in multiple samples across the entire genome: the outer circle in the diagram represents the chromosome, and the inner circle indicates the frequency of off-target and integration events with event density. High-frequency areas are clearly highlighted, facilitating rapid identification and analysis of potential high-risk areas.

[0159] Example 4

[0160] An in vivo detection system for simultaneous detection of CRISPR off-target sites and viral vector integration sites, comprising:

[0161] Data preprocessing module, which supports direct import of paired-end FASTQ files from the Illumina sequencing platform and automatically completes data quality control through built-in quality control tools, including adapter sequence removal and low-quality sequence filtering (Phred quality score Q < 30);

[0162] Demultiplex and UMI deduplication module, which automatically separates and identifies samples by reading index sequence files (Index files) and accurately extracts UMI sequences. It uses an efficient UMI clustering algorithm to remove redundant PCR amplification sequences to ensure high data accuracy.

[0163] A sequence alignment and off-target detection module, which uses the Smith-Waterman local alignment algorithm to achieve accurate alignment and quickly identify and mark potential off-target sites. This module automatically calculates the AviTag-seq score for each site and generates visualization charts, such as Circos plots and heat maps, to intuitively display the distribution of off-target events.

[0164] An AAV integration site analysis module, which adopts a high-efficiency algorithm to accurately identify the genomic integration site of an AAV vector, generates a detailed integration site analysis report by performing genomic annotation on the integration site (including gene symbol, start position, abundance, and other information);

[0165] An automated report generation module, which automatically integrates the analysis results into a standardized PDF or HTML report; the report content includes sample quality control information (quality control indicators, UMI deduplication rate), detailed lists of off-target sites and integration sites (site position, frequency, and related annotations), rich visual charts (Circos chart, chromosome heat map, gene region distribution chart, etc.), experimental parameter and analysis process information, ensuring the traceability and transparency of the results, system compatibility with Linux and Windows environments, support for batch analysis mode, and significantly improved data processing efficiency.

[0166] Through the above functional modules and the automated report generation system, the software realizes efficient, reliable, and industrial-grade AviTag-seq data analysis and report output, providing users with powerful data analysis capabilities and convenient operation experience.

[0167] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, but not to limit it, although the present application has been described in detail with reference to the above examples, those skilled in the art should understand that: the specific embodiments of the present application can still be modified or replaced equivalently without departing from the spirit and scope of the present application, any modification or equivalent replacement which does not depart from the spirit and scope of the present application should be covered within the protection scope of the claims of the present application.

Claims

1. An in vivo method for simultaneous detection of CRISPR off-target sites and viral vector integration sites, characterized in that: The following steps are involved: S1. Construct an AAV vector carrying an inverted terminal repeat sequence tag and co-deliver it with the CRISPR editing element to the cells or animals to be tested; S2. Collect cells or animal tissues, extract genomic DNA and fragment it; S3. After ligating the fragmented DNA to sequencing adapters, two rounds of PCR amplification were performed using first-round PCR primers containing unique molecular identifiers and second-round PCR primers containing universal ITR primers to obtain a sequencing library; S4. Use a paired-end high-throughput sequencing platform to obtain sequencing data; S5.UMI clustering removes repetitive amplified sequences, and then the Smith-Waterman local sequence alignment algorithm is used to accurately realign the structural fragments, precisely identifying off-target sites and AAV viral vector integration sites.

2. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S1, the serotype of the AAV vector is one of AAV2, AAV6, AAV8, AAV9 or an engineered variant thereof, and the genome length of the AAV vector is ≤4.0 kb.

3. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S1, the CRISPR editing element is one of SpCas9, eSpCas9, SaCas9, AsCas12a, LbCas12a, Prime Editor or Base Editor.

4. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S2, the fragmentation is to fragment into 300 to 800 bp.

5. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S4, the use of a double-end high-throughput sequencing platform to obtain sequencing data specifically includes: using the Illumina platform for 150bp double-end sequencing and the Nanopore platform LSK114 for long-read sequencing.

6. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S5, the UMI clustering to remove duplicate amplified sequences is specifically as follows: a. Data preprocessing The raw FASTQ files obtained from sequencing data were quality controlled to remove low-quality sequences and sequencing adapter sequences; b.UMI extraction and preliminary clustering: In sequencing data, each sequence starts with a UMI sequence of 8 to 12 bp in length, which is used to label the original template molecule for PCR amplification. First, the UMI tag of each sequence is extracted, and preliminary clustering and grouping are performed based on the UMI tag. c.UMI clustering deduplication algorithm: Based on the characteristics of the Illumina platform, a UMI clustering algorithm based on the edit distance threshold is used to cluster sequence UMIs. The threshold is optimized to an edit distance of 1, which means that only UMIs with a single base difference are allowed to be merged into the same cluster unit to effectively eliminate amplification bias caused by the PCR process. Each UMI cluster unit retains a representative sequence. d. Structural decoding analysis and double-end sequence splicing: The double-ended sequences after de-redundancy treatment were spliced ​​using Flash or PEAR tools. The overlapping region length after splicing was required to be ≥10bp and the mismatch rate was ≤0.1% to ensure the accuracy of the spliced ​​sequence. The complete sequence after splicing was used for the next analysis.

7. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S5, the Smith-Waterman local sequence alignment algorithm is used to accurately realign the structural fragments as follows: The sequences were aligned with the reference genome, and the optimization parameters were set as follows: match score +2, mismatch penalty -2, gap opening penalty -3, and gap extension penalty -1. After sequence alignment, the sequences were analyzed to determine whether they contained the characteristic structure of the AAV-ITR junction with the genome.

8. The in vivo detection method for synchronously detecting CRISPR off-target sites and viral vector integration sites according to claim 1, characterized in that In step S5, the precise identification of off-target sites and AAV viral vector integration sites is specifically as follows: A dual-index strategy was designed, with one end anchoring the AAV-ITR sequence features and the other end anchoring the genomic sequence features, to achieve accurate distinction and identification of off-target sites and AAV integration sites.

9. An in vivo detection system for simultaneously detecting CRISPR off-target sites and viral vector integration sites based on the in vivo detection method for simultaneously detecting CRISPR off-target sites and viral vector integration sites according to any one of claims 1 to 8, characterized in that: include: A data preprocessing module that supports direct import of paired-end FASTQ files from the Illumina sequencing platform and automatically performs data quality control using built-in quality control tools, including adapter sequence removal and low-quality sequence filtering. Demultiplex and UMI deduplication module, which automatically separates and identifies samples by reading index sequence files and accurately extracts UMI sequences. It also uses an efficient UMI clustering algorithm to remove redundant PCR amplification sequences to ensure high data accuracy. Sequence alignment and off-target detection module, which uses the Smith-Waterman local alignment algorithm to achieve accurate alignment and quickly identify and mark potential off-target sites. This module automatically calculates the AviTag-seq score for each site and generates a visual chart to intuitively display the distribution of off-target events. AAV integration site analysis module, which uses an efficient algorithm to accurately identify the genomic integration site of the AAV vector and generates a detailed integration site analysis report by annotating the integration site; An automated report generation module that automatically integrates analysis results into standardized PDF or HTML reports.

Images