AAV single-molecule sequencing library construction method, application and kit

Abstract

The invention discloses a construction method and application of an AAV single-molecule sequencing library and a kit, and belongs to the technical field of biological medicine detection. In order to solve the technical problem that ligase steric hindrance is caused by a secondary structure of an AAV inverted repeat (ITR), a Cas9-RNP compound specifically targeting an ITR stem region (Stem Region) is used for carrying out directional cutting on a single-chain genome, a closed-loop hairpin structure is removed, and a blunt-end double-chain DNA handle of 10-50 bp is reserved at the tail end. In cooperation with a forced dissociation step of the enzyme, the method eliminates connection steric hindrance and significantly improves the linker connection efficiency. According to the method, double-end cyclization is not needed, sequencing can be achieved only through single-end connection, and therefore a complete genome, a truncated body and a covalent head-to-tail concatemer can be captured without bias, and the method can be used for accurately measuring the physical packaging capacity limit of the AAV carrier.

Description

Technical Field

[0001] This invention belongs to the field of biomedical detection technology, and in particular relates to a method for constructing an AAV single-molecule sequencing library, its application, and a reagent kit. Background Technology

[0002] Recombinant adeno-associated virus (rAAV) vectors have become the primary delivery tool for in vivo gene therapy due to their high safety, broad tissue tropism, and ability to achieve long-term gene expression. With the approval of multiple rAAV gene therapies and the rapid growth of clinical pipelines, regulatory agencies in various countries (such as the FDA and NMPA) have imposed stricter requirements on product quality control. In particular, the detection of genome integrity and residual impurities has become a critical quality attribute (CQA) for batch release.

[0003] However, existing analytical methods have significant limitations. Traditional short-fragment methods such as qPCR, ddPCR, and next-generation sequencing (NGS) mainly rely on amplicon detection, which can only detect local sequences and cannot span the entire genome or accurately resolve the complex inverted repeat (ITR) structures at both ends. These ITR sequences are prone to forming highly stable T-shaped hairpin secondary structures, leading to steric hindrance in enzymatic reactions (such as ligases), resulting in low efficiency in constructing long-read sequencing libraries and high sample requirements.

[0004] Although single-molecule long-read sequencing technologies such as PacBio and Nanopore have been introduced to overcome the above problems, existing commercial solutions still suffer from detection blind spots and selection bias: PacBio SMRTbell library preparation relies on hairpin ligation at both ends to form a closed-loop template, resulting in the filtering of molecules with damaged ends or atypical structures, producing a "survivorship bias," which overestimates vector integrity and masks the true truncation ratio. Conventional Nanopore methods rely on natural annealing to form double strands, but intact complementary strands anneal more easily than truncated strands, causing enrichment bias. However, existing nanopore Cas9 targeted sequencing technologies (such as nCATS) are primarily designed for enriching long genome fragments and are not optimized for the extreme secondary structures of AAV ITRs. Two key physical obstacles remain unresolved in existing technologies: 1. End dilemma caused by cutting: Conventional designs often avoid the high GC region of the ITR, resulting in an excessively long residual ITR stem and incomplete opening of secondary structures; or the blunt ends produced by cutting are too short (<10 bp), which cannot provide the footprint required by T4 DNA ligase.

[0005] 2. Enzyme Retention: Cas9 exhibits an extremely slow turnover rate after cleavage, anchoring tightly to DNA ends and physically blocking the ligation of sequencing adapters. Current procedures lack a targeted forced dissociation step, resulting in extremely low ligation efficiency (<1%) in AAV samples.

[0006] Therefore, a specific structural remodeling process is urgently needed that can precisely eliminate the secondary structure of the ITR, retain a suitable length of connecting handle, and overcome the spatial resistance of the Cas9 itself. Summary of the Invention

[0007] The purpose of this invention is to provide a method, application, and kit for constructing AAV single-molecule sequencing libraries, thereby solving the technical problems in existing AAV sequencing library construction technologies, such as low ligation efficiency due to the complex secondary structure of ITR, large sample requirements, and terminal selection bias caused by doubly circularized or annealed dependence.

[0008] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for constructing an adeno-associated virus (AAV) single-molecule sequencing library, wherein the method utilizes the CRISPR-Cas9 ribonucleoprotein (RNP) complex to specifically remodel the secondary structure of the AAV ITR to eliminate steric hindrance of the ligase, and the method includes the following steps: Step S1: Extract and purify single-stranded DNA (ssDNA) genome from AAV virus particles; Step S2: The ssDNA genome is directionally cleaved using the Cas9-RNP complex; wherein the Cas9-RNP complex contains the Cas9 protein and sgRNA that specifically targets the stem region of the AAV ITR hairpin structure; the cleavage removes the closed-loop hairpin structure of the ITR and retains a 10-50 bp blunt-ended double-stranded DNA handle at the end as a ligase binding site; Step S3: The product of step S2 is subjected to protease treatment and heat denaturation to forcibly dissociate the Cas9 protein bound to the DNA ends and expose the blunt-ended double-stranded DNA handle. Step S4: The blunt ends released in step S3 are repaired and dA tails are added. Under the action of T4 DNA ligase, the sequencing adapter is connected to the retained blunt-end double-stranded DNA handle. Step S5: Purify the ligation product and load it into a nanopore sequencing chip for single-molecule sequencing.

[0009] Preferably, in step S2, the RNP complex comprises a Cas9 protein and a specific sgRNA that targets the AAV inverted repeat sequence ITR; The sgRNA guides Cas9 to perform double-strand cutting at a specific site inside the hairpin structure of the ITR, reshaping the T-shaped hairpin structure of the ITR into a double-stranded DNA structure with blunt ends. The double-stranded DNA with blunt ends retains a short double-stranded region of 10-50 bp at the ends.

[0010] Preferably, the ligation step does not depend on the circularization of both ends of the DNA molecule, such that either truncated DNA containing at least one blunt-ended double-stranded DNA handle or full-length genomic DNA can be ligated to the sequencing adapter and sequenced.

[0011] Preferably, the target sequence of the sgRNA is located in the stem region of the AAV ITR hairpin structure, and the target sequence is selected from the following sequences: SEQ ID NO.1:GAGCGAGCGAGCGCGCAGAG; SEQ ID NO.2: GCGCTCGCTCGCTCACTG; Or a variant sequence that has at least 90% identity with the above sequence and can achieve equivalent ITR cutting and structural opening.

[0012] Preferably, the method further includes the step of identifying covalent head-to-tail tandem structures in AAV vectors: Step A (structural remodeling sequencing): Perform the steps described in claim 1, and use Cas9-RNP to specifically cleave the ITR connection nodes between AAV genomes to obtain the first sequencing dataset; Step B (Conformation-preserving sequencing): Take AAV genomes from the same batch, without performing the Cas9-RNP treatment, and directly perform adapter ligation and sequencing after only heat denaturation and renaturation treatment to preserve the intermolecular or intramolecular covalent linkage structure and obtain the second sequencing dataset. Step C (Computer-aided feature extraction): Compare the first sequencing dataset and the second sequencing dataset in the data processing system; identify reads that are detected in the second dataset but have significantly reduced abundance or are missing in the first dataset, and whose length is an integer multiple of the designed length of the monogenome, as covalent head-to-tail tandem sequences.

[0013] Preferably, the method is further used to determine the physical packaging capacity limit of a specific AAV capping, the steps of which include: Construct a series of AAV vector libraries with increasing insertion fragment lengths; The library was sequenced using the method described in claim 1, and the genome integrity index of each vector was calculated. The vector length threshold at which the genome integrity index experiences a nonlinear drop is determined as the physical packaging capacity limit of the AAV capsid.

[0014] Preferably, the method is further used for the quantitative detection of residual DNA impurities, and the steps include: A composite reference system was constructed, comprising the AAV vector sequence, plasmid backbone sequence, and host genome sequence; Masking is performed on the ITR region and promoter region in the composite reference system that are homologous to the AAV vector, replacing homologous bases with degenerate bases or placeholders. The sequencing reads were aligned to a masked composite reference frame, and the content of residual plasmid backbone impurities was quantitatively calculated based on the specific matching reads.

[0015] Secondly, the present invention provides a kit for constructing AAV single-molecule sequencing libraries, comprising: a) Structural remodeling component: contains Cas9 nuclease and at least one sgRNA, wherein the sgRNA specifically targets the stem region of the AAVITR hairpin structure and its cleavage site is configured to generate a 10-50 bp blunt-ended double-stranded DNA handle at the end of the ITR. b) Enzyme dissociation component: Contains proteinase K and denaturation buffer, used to forcibly dissociate the Cas9 complex from the DNA ends; c) Sequencing adapter components: Contains adapters adapted to nanopore sequencing platforms.

[0016] Preferably, the sgRNA is designed for the conserved sequence of AAV2 ITR, thereby enabling universal structural remodeling for different serotype AAV vectors containing AAV2 ITR, including AAV8, AAV9, or AAV-DJ. The target sequence of the sgRNA is located in the stem region of the AAV ITR hairpin structure, and the target sequence is selected from the following sequences: SEQ ID NO.1: GAGCGAGCGAGCGCGCAGAG; SEQ ID NO.2: GCGCTCGCTCGCTCACTG; Or a variant sequence that has at least 90% identity with the above sequence and can achieve equivalent ITR cutting and structural opening.

[0017] Thirdly, the present invention provides the application of the above-described method in the quality control of AAV gene therapy vectors, characterized in that it is used for one or more of the following: Unbiased assessment of AAV vector genome integrity; Identification and quantification of covalent head-to-tail tandem structures; Precisely determine the upper limit of the physical packaging capacity of a specific AAV cap, i.e., the packaging cliff; Highly sensitive quantitative detection of residual plasmid backbone, helper plasmids and host DNA impurities.

[0018] The beneficial effects of this invention are as follows: 1. This invention eliminates spatial steric hindrance, resulting in a qualitative leap in connection efficiency. This invention utilizes Cas9-RNP to transform difficult-to-ligate ITR hairpins into standard blunt-terminated dsDNA substrates. Unexpectedly, this structural remodeling step resulted in an explosive increase in ligation efficiency: experimental data show that the ligation efficiency after Cas9-RNP treatment was an astonishing 7-10 times (i.e., 700%-1000%) compared to the untreated group. Typical optimization of enzymatic reaction conditions only yields a 20-50% efficiency improvement, while the order-of-magnitude improvement achieved by this invention confirms that the steric hindrance effect of ITRs is far more severe than anticipated by existing technologies, and that the solution of this invention has extremely significant and non-obvious technical effects. This makes it possible to obtain high-depth sequencing data with extremely low DNA input (down to the nanogram level).

[0019] 2. This invention eliminates selection bias and achieves "blind zone-free" detection. This invention employs a "single-end ligation is effective" strategy, overcoming the "survivor bias" of PacBio technology, which requires both ends to be intact for circularization. Both full-length genomes and truncations with single-end damage have equal opportunities for ligation and sequencing, thus accurately reflecting the full heterogeneity of viral vectors.

[0020] 3. This invention is the first to accurately identify "covalent head-to-tail tandem". The dual-mode analysis strategy (RNP mode vs. annealing mode) proposed in this invention is the first to distinguish between physically annealed products and biologically covalent head-to-tail tandems. Experiments have confirmed that a large number of covalent dimers or polymers, which are ignored by existing technologies, exist in vectors with a length of less than 3.0 kb. The detection sensitivity of this invention for such structures is significantly better than that of the PacBio HiFi platform.

[0021] 4. The present invention provides a general method for accurately defining the physical packaging ciff. Based on unbiased integrity data, this invention provides a universal "molecular ruler" capable of independently determining the specific physical packaging capacity upper limit for different AAV serotypes (such as AAV2, AAV5, AAV8, AAV9, and novel artificial capsids). For example, in validation experiments targeting AAV8, this method precisely captured the precipitous drop in genome integrity at 5.0–5.2 kb. This measurement capability is independent of specific serotype assumptions, providing accurate and measurable physical boundary parameters for the customized design of various AAV vectors. Attached Figure Description

[0022] Figure 1 The diagram shows a comparison of the library construction principles (a) and segmentation (b) between the method of this invention (AviNP-seq) and the prior art (PacBio SMRTbell); Figure 2 This is a graph showing the effect of the Cas9-RNP processing step in this invention on improving sequencing adapter ligation efficiency. Figure 3 This is a panoramic view of the genome integrity of AAV vectors of different lengths as determined by the method of this invention. Figure 4 The results of identifying covalent tandem molecules in low molecular weight carriers using the dual-mode analysis strategy of this invention are shown in the figure. Figure 5 This is a comparison chart of the detection rate of tandem bodies between the method of this invention and the existing PacBio HiFi sequencing technology; Figure 6 This is a performance verification diagram of the method of the present invention for the quantitative analysis of residual DNA impurities. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present invention, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The following embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0024] Definitions of terms used in this invention: ITR structural remodeling refers to the process of specifically altering the physical conformation of the terminal inverted repeat (ITR) sequence of the adeno-associated virus (AAV) genome using enzymatic biochemical reactions. In its natural state, the ITR region of the single-stranded AAV genome forms a high-energy, stable T-shaped hairpin structure through intramolecular base pairing. This structure constitutes significant steric hindrance to subsequent enzymatic reactions (especially ligase reactions). Specifically, the "structural remodeling" described in this invention refers to utilizing the R-loop formation ability, strand invasion mechanism, and local double-strand dissociation ability generated when the CRISPR-Cas9 ribonucleoprotein (RNP) complex binds to the target sequence to open the hairpin structure. Then, its endonuclease activity is used to perform double-strand cleavage at specific sites within the ITR, thereby transforming the originally closed, folded T-shaped hairpin into an open, blunt-end dsDNA structure that can serve as a standard substrate for ligation reactions. This process is not simply about cutting out sequences; its core purpose is to eliminate steric hindrance, making ends that would otherwise be unable to be captured by ligases "ligation-ready".

[0025] One-end-sufficient: This refers to a strategy for constructing and initiating single-molecule sequencing libraries. Unlike the existing PacBio SMRTbell library construction strategy (which requires simultaneous successful ligation of dumbbell-shaped adapters at both the 5' and 3' ends of the DNA molecule to form a closed loop), the "one-end-sufficient" strategy employed in this invention means that during library construction, it is not necessary to simultaneously ligate or circularize both ends of the DNA molecule; as long as either end of the DNA molecule (5' or 3') exposes a ligable end (such as a blunt end after Cas9 cutting) and successfully ligates a sequencing adapter (especially an adapter for motor proteins), the molecule can be captured by the nanopore sequencer and sequencing can begin. The essential advantage of this strategy is that it eliminates the mandatory requirement of "double-end integrity," thereby enabling the capture and sequencing of molecules with damaged, truncated, or structurally incomplete ends, fundamentally avoiding "survivorship bias."

[0026] Packaging cliff: This refers to the phenomenon where, under the physical constraints of the AAV capsid, the viral genome length exceeds a certain critical threshold, resulting in a precipitous drop in genome integrity. The specific value of this critical threshold depends on the capsid physical properties of a particular AAV serotype (e.g., 5.0-5.2 kb in the embodiments of this invention for AAV8). The core advantage of the method described in this invention is its ability to accurately locate this critical threshold for any serotype using unbiased sequencing data, without relying on pre-set theoretical values.

[0027] Unbiased Denominator: When calculating genome integrity (number of complete molecules / total number of molecules), the "total number of molecules" is the denominator. Traditional methods rely on annealing or biterminal circularization, resulting in the loss of a large number of truncated molecules in the denominator, leading to an underestimation of the denominator and an artificially inflated integrity value. This invention utilizes Cas9-RNP processing to capture all molecules, providing a true complete set including truncated molecules as the denominator, thereby correcting the bias in integrity calculation.

[0028] Example 1 Assembly of the CRISPR-Cas9 RNP complex and sgRNA design: • Cas9 protein: This invention uses high-fidelity, purified recombinant Streptococcus pyogenes Cas9 protein (SpCas9), namely Alt-R Sp Cas9 Nuclease V3 (IDT, Cat# 1081058), which must be free of nuclease contamination that would lead to nonspecific degradation.

[0029] • sgRNA design principles: The sgRNA of this invention is designed not for specific transgene internal sequences, but for the ITR sequence, a universal element of the AAV vector.

[0030] To ensure the universal applicability of the method, the target sequence of the present invention is located in the stem or non-hairpin region of the ITR hairpin structure, and the sequence is highly conserved in AAV2 ITR.

[0031] The sequence obtained by this invention is as follows: Sequence 1 (sgITR-16): GAGCGAGCGAGCGCGCAGAG (SEQ ID NO. 1). This sequence guides Cas9 to cleave near the end of the ITR, preserving a short double-stranded blunt end of approximately 16 bp.

[0032] Sequence 2 (sgITR-34): GCGCTCGCGCTCTCACTG (SEQ ID NO. 2). This sequence guides Cas9 to cut further inside the ITR, preserving a short double-stranded blunt end of approximately 34 bp.

[0033] The composite of the present invention is assembled as follows: In order to form a functional ribonucleoprotein (RNP) complex, the present invention mixes Cas9 protein with sgRNA in a molar ratio of 1:2 to 1:5 (the present invention chooses 1:2).

[0034] Specifically, 1 µL of Cas9 protein (62 µM) was mixed with 4 µL of sgRNA (30 µM) and incubated at 37°C for 10 minutes to complete assembly.

[0035] Compared to plasmid transfection to express Cas9, the present invention uses the RNP complex directly to achieve instantaneous and highly active cleavage, and avoids the introduction of exogenous plasmid DNA that interferes with the sequencing background.

[0036] This invention, based on a theoretical model of steric hindrance and ligase binding requirements, screened and evaluated different target regions. To determine the optimal structural remodeling site, this invention presented a logical table of three sets of sgRNA sites targeting different regions of AAV ITR: (see Table 1): Table 1. Logic table for screening three groups of sgRNA sites targeting different regions of AAV ITR

[0037] Group A (Loop region): Due to the extremely high base stacking force and steric hindrance at the hairpin end of the ITR, the activity of Cas9 helicase is limited, resulting in a messy distribution of cleavage sites (Smear), making it difficult to form an effective library initiation end.

[0038] Group B (RBE region): The cleavage site is too close to the end, resulting in a blunt-ended double-stranded region that is insufficient to support the stable binding of T4 ligase, thus hindering the kinetics of the adapter ligation reaction.

[0039] Group C (Stem region of this invention): Targeting the ITR stem (SEQ ID NO: 1, sgRNA-Stem). The hairpin structure was removed by cleavage, and a 16 bp blunt-ended double-stranded handle was retained at the end. This length perfectly matches the binding groove of the T4 ligase, achieving the highest ligation efficiency.

[0040] Conclusion: Not all cuts targeting the ITR are effective; only stem cuts that retain 10-50 bp of the handle can achieve efficient library construction.

[0041] The above results demonstrate that not all sgRNAs designed for ITR sequences can achieve the technical effects of this invention. Because the AAV ITR region forms a highly stable T-shaped hairpin secondary structure (extremely low ΔG) and has a very high GC content, sgRNAs recommended by conventional bioinformatics design tools often fail to effectively guide Cas9 protein into the double strand, resulting in extremely low cleavage efficiency (<10%).

[0042] In addition, during the research and development process, this invention tested multiple sgRNAs targeting different regions of the ITR (such as RBE site, TRS site, and hairpin tip loop region). It was found that although some sites could be cleaved, the resulting residual double strands were too short (<10 bp), resulting in unstable ligation of subsequent sequencing adapters; or the residual structures were too long, failing to completely eliminate spatial steric hindrance.

[0043] Through systematic screening, this invention identified the specific "golden target" sequences (SEQ ID NO: 1 and 2). These sequences specifically target a specific stem position of the ITR hairpin structure, where cleavage achieves an optimal balance between steric hindrance release and end length: fully opening the secondary structure while precisely preserving the optimal blunt end length (16 bp or 34 bp) suitable for T4 ligase binding. Those skilled in the art will understand that the T4 DNA ligase footprint typically requires at least 10-12 bp of double-stranded support. Excessively long preserved sequences (e.g., exceeding 50 bp) may cause the ITR secondary structure to refold, again hindering enzyme binding. Therefore, the 10-50 bp range defined in this invention is a reasonable range based on enzymatic properties and thermodynamic stability. Comparative experiments show that library construction efficiency using the preferred sgRNAs of this invention is significantly better than sgRNAs targeting other regions of the ITR, demonstrating unexpected technical advantages.

[0044] Sequencing Adapter This invention selects a ligation sequencing kit adapted to the nanopore sequencing platform, namely Oxford Nanopore Technologies (ONT)'s Native Barcoding Kit96 V14 (SQK-NBD114.96).

[0045] The kit contains a "motor adapter (such as a native adapter (NA))" with one end being a blunt end or a dT overhang, which can be connected to an AAV genome that has undergone end repair and dA tailing; the other end is loaded with a helicase (motor protein) to pull single strands of DNA through nanoporous proteins.

[0046] Example 2 Cas9-RNP mediated library construction process This embodiment will describe in detail the standard operating procedure (SOP) for eliminating ITR steric hindrance and constructing single-molecule sequencing libraries using Cas9-RNP. This procedure is the core step in achieving the "blind zone-free" detection of this invention.

[0047] Step 1: AAV single-stranded genome extraction (ssDNA Extraction) (1) Take an appropriate amount of purified AAV virus solution (the titer used in this example is 1E12VG / mL) and treat it with DNase I to remove free DNA outside the capsid.

[0048] (2) Add DNA extraction buffer and proteinase K. The genomic DNA extraction buffer consists of: 100 mM sodium chloride, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, and 0.5% Tween-20. Before use, add proteinase K to the buffer at a final concentration (0.1 mg / mL). Incubate at 56°C for 1 hour to destroy the viral capsid protein and release single-stranded DNA (ssDNA).

[0049] (3) Purify ssDNA using magnetic beads (AMPure XP beads are selected in this invention).

[0050] Step 2: ITR Structure Remodeling and End Release (RNP) Reaction system: Add 1 µg of purified ssDNA, 5 µL of pre-assembled Cas9-RNP complex (Cas9 to sgRNA molar ratio 1:2, Cas9 and sgRNA molars 60 pmol and 120 pmol respectively), and 2 µL of 10× NEBuffer r3.1 to a 20 µL reaction system. Incubate at 37°C for 120 minutes. The extended incubation time ensures that Cas9 can capture high-GC ITR structures in a dynamic breathing state.

[0051] Technical Principle Explanation: Conventional restriction endonucleases cannot recognize the hairpin structure of the ITR in single-stranded AAV. After binding to sgRNA, the Cas9 protein acts like a wedge, actively invading and unwinding the high-GC hairpin structure of the ITR, and then precisely cleaving the blunt ends of the double strands under the guidance of the sgRNA. The extended incubation time (2 hours) ensures that all tightly folded ITR structures are fully opened and cleaved.

[0052] Product state: After the reaction is complete, the originally closed ITR hairpins that hindered ligation are converted into open double-stranded DNA (dsDNA) with blunt ends.

[0053] Step 3: Enzyme inactivation and removal Add 0.2 µL of proteinase K (1%) to the reaction solution again, and treat at 37°C for 15 minutes, then at 95°C for 5 minutes to detach protein fragments and transiently denature the DNA. This step exposes the blunt ends of the DNA that were previously masked by Cas9. The DNA is then purified using magnetic beads (AMPure XP beads were chosen in this invention).

[0054] Technical principle explanation: After cleavage, the Cas9 protein tends to bind tightly to the DNA end and does not detach (slow turnover), which physically prevents subsequent ligases from approaching. The Cas9 protein must be forcibly degraded by proteinase K to release the naked blunt ends of the DNA, allowing adapter ligation to occur.

[0055] Step 4: End-Prep & Ligation (1) End repair: Add the end repair enzyme mixture (End Prep EnzymeMix) and buffer directly to the product of step 3. For details, use the Native Barcoding Kit 96 V14 protocol provided by the commercial Oxford Nanopore Kit ONT Ligation Sequencing Kit V14 (SQK-LSK114). React at 20°C for 5 minutes and then at 65°C for 5 minutes.

[0056] Technical principle explanation: This step adds an "A" base to the blunt end of the Cas9.

[0057] (2) Adapter ligation: Add nanopore sequencing adapter (Native Adapter (NA) in this invention) and high concentration of T4 DNA ligase (Quick T4 Ligase). For details, refer to the Native Barcoding Kit 96 V14 protocol provided by the commercially available Oxford Nanopore Kit ONTLigation Sequencing Kit V14 (SQK-LSK114) and incubate at room temperature for 30 minutes.

[0058] Step 5: Purification and Sequencing (1) Use 0.4x magnetic beads to purify the ligation product to remove unligated adapters and enzyme residues; (2) Load the library onto the nanopore sequencing chip (Flow Cell R10.4.1) for sequencing and collect raw electrical signal data.

[0059] To verify whether the Cas9-RNP was precisely cut as designed, we compared the sequencing reads back to the reference genome and visualized them in IGV software. Specifically, the Fast5 file obtained after sequencing was converted to Fastq using Dorado v7.6.7 software, and then compared back to the vector reference sequence using Minimap2. The start and end sites of the reads were observed in IGV (Integrative GenomicsViewer). If the cut was precise, the reads should show a clean, abrupt break at the specified position on the sgRNA, rather than random fragmentation, as shown in the results. Figure 1 As shown.

[0060] from Figure 1 As can be seen, after treatment with specifically designed sgRNAs (such as RNP-16 and RNP-34), the sequencing coverage exhibits extremely neat and sharp edges at the expected cleavage sites (within the ITR), with arrows clearly indicating the cleavage locations. This demonstrates that Cas9-RNP not only opens the structure but also achieves precise "molecular surgery" at single nucleotide resolution, providing clear boundary features for subsequent bioinformatics analysis.

[0061] Example 3 The method of this invention significantly improves ligation efficiency and sequencing yield. This embodiment aims to quantify the specific contribution of "ITR structure reshaping" to library construction efficiency, and to demonstrate the "unexpected technical effects" brought about by this step.

[0062] The experimental design for this embodiment is as follows: Control group (annealing only, No RNP): DNA was only denatured at 95°C and then slowly cooled to renature, relying on partial double-stranded regions for ligation.

[0063] Experimental group (Cas9-RNP group): DNA was processed by Cas9-RNP cutting as described in Example 1.

[0064] To achieve absolute quantification, this invention incorporates an equal amount (0.1% w / w) of linear double-stranded DNA fragment (dsPCR spike-in) as an internal control in both groups of samples, and conducts experiments with the ligation efficiency of the internal control being 100%.

[0065] Experimental procedure: To achieve a quantitative and normalized assessment of AAV genome adapter ligation efficiency, this invention introduces a 4750 bp exogenous double-stranded PCR product (dsDNA Spike-in) as a standardized internal control. Before the end-prep reaction in step 3, 0.1% (by mass) of the aforementioned internal control DNA is precisely added to the AAV genome extraction product system (e.g., if the total AAV DNA is 1000 ng, then 10 ng of internal control is added). This internal control sequence has no homology with either the AAV genome or the host genome, ensuring the uniqueness of the bioinformatics back-match. By calculating the ratio of reads aligned to the AAV sequence and the internal control sequence in the offline data, and eliminating molecular length differences (i.e., calculating: [AAV Reads / 4.75 kb] / [Internal Control Reads / 4.75 kb]), the normalized ligation efficiency, excluding pooling state interference, can be obtained. Experimental results show that the normalization efficiency of the Cas9-RNP pretreated group is significantly improved compared with the untreated group, providing a reliable quantitative criterion for blind spot-free characterization of the AAV genome.

[0066] By comparing the ratio of intrinsic reads to virus reads, this invention calculates the normalized connection efficiency, and the experimental results are as follows: Figure 2 As shown.

[0067] from Figure 2 As can be seen, the adapter ligation efficiency of the Cas9-RNP treatment group reached 7-10 times (700%-1000%) that of the control group (P<0.0001), far exceeding the expected range of conventional biochemical reaction optimization. This means that with the same amount of DNA input, the method of this invention can convert more viral molecules into an effective library.

[0068] Meanwhile, it can be seen that in actual sequencing output, the number of effective viral reads obtained by the RNP group is 7-8 times that of the control group.

[0069] The above results demonstrate that the secondary structure of AAV ITR constitutes a strong, underdeveloped steric hindrance for ligases. This invention, by "forcing" the opening of this structure using Cas9-RNP, achieves a qualitative leap from "difficult ligation" to "efficient ligation." This unexpected and significant technical effect powerfully demonstrates the non-obviousness of this method.

[0070] Example 4 This embodiment provides a precise definition of the AAV packaging limit. In this embodiment, using the unbiased sequencing library constructed according to the present invention, the physical packaging boundary of the AAV capsid is accurately depicted for the first time. This achieves an important re-understanding of the biological characteristics of AAV.

[0071] This invention constructs a series of AAV vector panels of increasing length, with the insert fragment gradually increasing from 1.5 kb to 6.5 kb. These vectors are then sequenced using the AviNP-seq method described in Example 1.

[0072] The results obtained are as follows Figure 3 As shown, the complete genome (Intact Genome) is defined as a read whose alignment length on the reference sequence covers more than 95% of the distance from 5' ITR to 3' ITR.

[0073] from Figure 3 The results show that as the vector design length increases, genome integrity exhibits a specific pattern of change: (1) Within the range of <5.0 kb, the integrity remains at a high level; (2) When the length reaches the range of 5.1 - 5.2 kb, the integrity curve drops sharply (The Packaging Cliff). (3) When the length is >5.2 kb, the integrity drops to almost zero. At this time, the main product detected by sequencing is no longer the full-length genome, but a truncated fragment with a length of about 3.0 kb.

[0074] The results above demonstrate that only by relying on the unbiased library construction method of this invention (which captures all truncated sections) can such a clear and accurate "packaging cliff" curve be obtained. If PacBio (which filters out truncated sections) or ordinary Nanopore (which tends to be full-length) is used, the position of this cliff will be artificially delayed or obscured.

[0075] Although this embodiment demonstrates that the packaging limit of the AAV8 serotype is in the 5.0-5.2 kb range, it should be understood that different AAV serotypes (such as AAV5 or artificially evolved capsids) may have different internal volumes due to slight differences in capsid protein structure, thus exhibiting different packaging cliff positions (e.g., AAV5 may have a slightly larger packaging capacity). Therefore, a significant contribution of this invention is to provide a standardized assay that can accurately determine the specific packaging hard upper limit for any given AAV serotype, eliminating capacity misjudgments caused by detection technology bias. This data has guiding significance for the design of gene therapy drug delivery systems.

[0076] Example 5 This embodiment provides a method for discovering covalent conjoints using a dual-mode strategy. (1) For AAV vectors with shorter length (<3 kb), this invention proposes a unique “Dual-Mode” analysis strategy to distinguish and identify covalent tandem vectors.

[0077] The specific operation process of this dual-mode strategy is as follows: Mode A (Annealing Mode): Without Cas9, the molecules are joined only after annealing. This mode preserves the original connection state of the molecules.

[0078] Mode B (RNP Mode): Adds Cas9 to cut the ITR. This mode will sever the ITR junctions that connect molecules.

[0079] The sequencing results obtained from mode A are as follows Figure 4 As shown in A, the sequencing results obtained from mode B are as follows: Figure 4 As shown in B.

[0080] (2) The same sample was sent to PacBio HiFi sequencing, and the results were as follows: Figure 5 As shown.

[0081] from Figure 4 The results in A show that for a 1.5 kb vector, a large number of reads with lengths of 3.0 kb (2 times the length, Dimer) and even 4.5 kb (3 times the length, Trimer) were observed.

[0082] from Figure 4 The results of B show that the long segment signal completely disappeared, and the data reverted to a single-length Monomer.

[0083] The results above show that dimers formed solely by physical adsorption or annealing should dissociate during library construction in Mode A (which involves denaturation steps). However, we observed stable dimers in Mode A, while they disappeared in Mode B (ITR cleavage). This definitively proves that these polyploids, covalently linked by ITRs in "head-to-tail tandems," are biologically real structures.

[0084] from Figure 5 The results show that the proportion of dimers detected by PacBio HiFi is significantly lower than that of Mode A of this invention (P=0.0016). This is because the two ends of covalent tandem polymers often have complex ITR structures, making it difficult to simultaneously form SMRTbell dumbbell rings, resulting in severe false negatives for PacBio. This result demonstrates that the dual-mode strategy of this invention fills this detection blind spot.

[0085] Example 6 This embodiment provides a "masking" analysis method for trace impurities. In addition to analyzing the viral genome itself, this invention also provides a set of supporting bioinformatics workflows for highly sensitive detection of residual DNA impurities in the same sequencing data. The bioinformatics workflow is as follows: When constructing the composite reference genome, the plasmid backbone sequence was aligned with the AAV vector sequence using the BLASTN algorithm. All regions with homology >90% and length >50 bp (mainly ITR and promoter elements) were identified.

[0086] In the skeleton reference sequence, the identified homologous region sequence is replaced with consecutive "N" characters (HardMasking).

[0087] During read alignment, any reads that align to the “N” region are discarded, and only reads that align to backbone-specific regions (such as the AmpR gene and the Ori origin of replication) are retained.

[0088] Construct a composite reference system that includes "AAV vector + plasmid backbone + helper plasmid (Helper / RepCap) + host genome (Host / Human)".

[0089] Since AAV vector sequences are usually derived from plasmids, there are a large number of homologous sequences (such as ITRs) between the two. In order to prevent reads from being incorrectly aligned to the backbone, homologous regions in the composite reference frame must be masked (i.e., homologous regions are replaced with "N"), retaining only the characteristic sequences unique to each component.

[0090] This invention involves artificially incorporating different proportions (0% - 5%) of linearized plasmid backbone DNA into a purified AAV library. By artificially incorporating different mass ratios (0.05%, 0.1%, 0.25%, 0.5%, 1%, 2%, and 5% w / w) of linearized 6.6kJ plasmid backbone DNA (with no homology to the AAV packaging plasmid, ensuring the uniqueness of the bioinformatics backtest) into purified AAV vector samples, the linearity and sensitivity of the AviNP-seq technology for detecting process-related impurities were verified. The results are as follows: Figure 6 As shown.

[0091] from Figure 6 It can be seen that the detected skeleton read ratio exhibits a very high linear correlation with the incorporation ratio (R0). 2 ≈ 0.956). Moreover, even at a doping ratio as low as 0.05%, this method can still stably detect skeleton reads (59 reads) with extremely low background noise.

[0092] The sensitivity of the above results (0.05%) fully meets the detection requirements of regulatory agencies in various countries for residual DNA impurities in gene therapy products. Therefore, this invention eliminates the need to design specific qPCR probes, enabling simultaneous vector integrity analysis and full-spectrum impurity quantification in a single sequencing run, greatly simplifying the batch release testing process.

Claims

1. A method for constructing an adeno-associated virus (AAV) single-molecule sequencing library, characterized in that, The method utilizes the CRISPR-Cas9 ribonucleoprotein (RNP) complex to specifically remodel the secondary structure of AAV ITR to eliminate steric hindrance of the ligase. The method includes the following steps: Step S1: Extract and purify single-stranded DNA (ssDNA) genome from AAV virus particles; Step S2: The ssDNA genome is directionally cleaved using the Cas9-RNP complex; wherein the Cas9-RNP complex contains the Cas9 protein and sgRNA that specifically targets the stem region of the AAV ITR hairpin structure; the cleavage removes the closed-loop hairpin structure of the ITR and retains a 10-50 bp blunt-ended double-stranded DNA handle at the end as a ligase binding site; Step S3: The product of step S2 is subjected to protease treatment and heat denaturation to forcibly dissociate the Cas9 protein bound to the DNA ends and expose the blunt-ended double-stranded DNA handle. Step S4: The blunt ends released in step S3 are repaired and dA tails are added. Under the action of T4 DNA ligase, the sequencing adapter is connected to the retained blunt-end double-stranded DNA handle. Step S5: Purify the ligation product and load it into a nanopore sequencing chip for single-molecule sequencing.

2. The method according to claim 1, characterized in that, In step S2, the RNP complex contains the Cas9 protein and a specific sgRNA that targets the AAV inverted repeat sequence ITR. The sgRNA guides Cas9 to perform double-strand cutting at a specific site inside the hairpin structure of the ITR, reshaping the T-shaped hairpin structure of the ITR into a double-stranded DNA structure with blunt ends. The double-stranded DNA with blunt ends retains a short double-stranded region of 10-50 bp at the ends.

3. The method according to claim 1, characterized in that, The ligation step does not depend on the circularization of both ends of the DNA molecule, so that either truncated DNA containing at least one blunt-ended double-stranded DNA handle or full-length genomic DNA can be ligated to the sequencing adapter and sequenced.

4. The method according to claim 3, characterized in that, The target sequence of the sgRNA is located in the stem region of the AAV ITR hairpin structure, and the target sequence is selected from the following sequences: SEQ ID NO.1:GAGCGAGCGAGCGCGCAGAG; SEQ ID NO.2: GCGCTCGCTCGCTCACTG; Or a variant sequence that has at least 90% identity with the above sequence and can achieve equivalent ITR cutting and structural opening.

5. The method according to claim 1, characterized in that, The method further includes the step of identifying covalent head-to-tail tandem structures in AAV vectors: Step A (structural remodeling sequencing): Perform the steps described in claim 1, and use Cas9-RNP to specifically cleave the ITR connection nodes between AAV genomes to obtain the first sequencing dataset; Step B (Conformation-preserving sequencing): Take AAV genomes from the same batch, without performing the Cas9-RNP treatment, and directly perform adapter ligation and sequencing after only heat denaturation and renaturation treatment to preserve the intermolecular or intramolecular covalent linkage structure and obtain the second sequencing dataset. Step C (Computer-aided feature extraction): Compare the first sequencing dataset and the second sequencing dataset in the data processing system; identify reads that are detected in the second dataset but have significantly reduced abundance or are missing in the first dataset, and whose length is an integer multiple of the designed length of the monogenome, as covalent head-to-tail tandem sequences.

6. The method according to claim 1, characterized in that, The method is further used to determine the physical packaging capacity limit of a specific AAV cap, the steps of which include: Construct a series of AAV vector libraries with increasing insertion fragment lengths; The library was sequenced using the method described in claim 1, and the genome integrity index of each vector was calculated. The vector length threshold at which the genome integrity index experiences a nonlinear drop is determined as the physical packaging capacity limit of the AAV capsid.

7. The method according to claim 1, characterized in that, The method is further used for the quantitative detection of residual DNA impurities, and the steps include: A composite reference system was constructed, comprising the AAV vector sequence, plasmid backbone sequence, and host genome sequence; Masking is performed on the ITR region and promoter region in the composite reference system that are homologous to the AAV vector, replacing homologous bases with degenerate bases or placeholders. The sequencing reads were aligned to a masked composite reference frame, and the content of residual plasmid backbone impurities was quantitatively calculated based on the specific matching reads.

8. A kit for constructing AAV single-molecule sequencing libraries, characterized in that, include: a) Structural remodeling component: contains Cas9 nuclease and at least one sgRNA, said sgRNA specifically targeting the stem region of the AAV ITR hairpin structure, and its cleavage site is configured to generate a 10-50 bp blunt-ended double-stranded DNA handle at the end of the ITR. b) Enzyme dissociation component: Contains proteinase K and denaturation buffer, used to forcibly dissociate the Cas9 complex from the DNA ends; c) Sequencing adapter components: Contains adapters adapted to nanopore sequencing platforms.

9. The reagent kit according to claim 8, characterized in that, The sgRNA is designed for the conserved sequence of AAV2 ITR, thereby enabling universal structural remodeling for different serotype AAV vectors containing AAV2 ITR, including AAV8, AAV9, or AAV-DJ. The target sequence of the sgRNA is located in the stem region of the AAV ITR hairpin structure, and the target sequence is selected from the following sequences: SEQ ID NO.1: GAGCGAGCGAGCGCGCAGAG; SEQ ID NO.2: GCGCTCGCTCGCTCACTG; Or a variant sequence that has at least 90% identity with the above sequence and can achieve equivalent ITR cutting and structural opening.

10. The application of the method according to any one of claims 1-7 in the quality control of AAV gene therapy vectors, characterized in that, Used for one or more of the following: Unbiased assessment of AAV vector genome integrity; Identification and quantification of covalent head-to-tail tandem structures; Precisely determine the upper limit of the physical packaging capacity of a specific AAV cap, i.e., the packaging cliff; Highly sensitive quantitative detection of residual plasmid backbone, helper plasmids and host DNA impurities.

Images