(W1130-03-15) NGS-Based Platform for Detecting Lentiviral Insertion Sites and Genomic Integration

Purpose: Lentiviral vectors are widely used in gene therapy and research, necessitating precise integration site characterization to ensure safety and efficacy. Integration Site Analysis (ISA) is crucial for assessing oncogenic risk and clonal expansion, as required by the Food and Drug Administration (FDA) and European Medicines Agency (EMA). To meet the demand for a reliable, high-throughput ISA assay, an NGS-based platform is needed for accurate lentiviral integration site mapping, supporting regulatory compliance and therapeutic development.

Methods: The assay was evaluated across multiple biological matrices, including human whole blood, peripheral blood mononuclear cells (PBMCs), and in vitro cultured cell lines. Assay performance metrics—including accuracy, precision, sensitivity, specificity, and stability—were assessed by the College of American Pathologists (CAP) guidelines. DNA Extraction and Library Preparation: Genomic DNA was extracted from whole blood, PBMCs, and cell lines using standard protocols. The extracted DNA was sheared to an average fragment size of approximately 300 bp, followed by ligation of asymmetric adapters to enable selective amplification. The Library preparation consisted of two rounds of PCR amplification interspersed with two rounds of size selection to optimize fragment distribution and reduce sequencing bias. Library quality was assessed using the Agilent TapeStation system prior to sequencing. Sequencing: High-quality sequencing libraries were prepared and processed on the Illumina NextSeq platform according to the manufacturer's standard protocols. Each sample yielded more than 10 million raw reads, providing high-accuracy, genome-wide short-read data for downstream analysis. To validate insert detection results, targeted PCR amplification was performed using primers flanking the predicted insertion sites. PCR products were purified and subjected to Sanger sequencing using standard procedures. The resulting chromatograms were aligned to the reference and insert sequences to verify integration sites, assess junction accuracy, and confirm the absence of indels or structural rearrangements. Representative integration events identified by NGS were selected for validation across samples. Data Analysis: NGS data were processed in three main steps; 1) Preprocessing: Raw sequencing data were demultiplexed, sequencing adapter elements were removed, and reads were subjected to quality control and filtering. Quality metrics were calculated to assess data integrity. 2) Sequence Alignment: Filtered high-quality reads were aligned to a composite reference genome comprising both the host genome and the exogenous sequence. Alignment metrics such as mapping rate and sequencing depth were calculated. 3) Integration Site Detection and Annotation: Based on the alignment results, integration sites were identified and quantified. Genomic features associated with integration loci were annotated to assess distribution characteristics.

Results: 1. Sample Library quality inspection result: TapeStation analysis confirmed that the libraries exhibited a single peak with a fragment size distribution between 400 and 600 bp, indicating appropriate library size and high quality for sequencing. 2. Sequencing quality Statistics: Sequencing quality control metrics for all samples are summarized in Table 1. All samples passed quality thresholds, with Q30 scores ≥ 75% and raw data volumes ≥ 1 Gb, and were deemed suitable for downstream analysis. 3. Sequence Alignment Result Statistics: After filtering the raw reads, clean reads were obtained and aligned to a composite reference genome comprising both the host and exogenous sequences. Alignment was performed using Bowtie2 with default parameters. The statistical alignment results are shown in Table 2. 4. Integration Site Detection: Custom in-house scripts were employed to identify breakpoints from the alignment results of each sample. Following quality filtering, integration breakpoints representing exogenous sequence insertions into the host genome were extracted. These integration sites were further annotated using the same in-house pipeline to provide information on gene names, gene functional regions, exon locations, and relevant database annotations. Sanger sequencing of positive control samples confirmed the accuracy of the detected integration events, showing 100% sequence identity with the corresponding reference regions on chromosomes 16 and 20. 5. Accuracy/Precision/Sensitivity/Specificity/Stability: The NGS-based platform demonstrated high accuracy, with results consistent with Sanger sequencing. Precision was confirmed through repeated measurements, ensuring reproducibility across replicates. The method achieved a sensitivity of 0.05%, enabling the detection of low-frequency insertion events, while specificity was validated by negative controls yielding no detectable signals. Additionally, the platform showed excellent stability, maintaining consistent lentiviral insertion site detection in whole blood and PBMC samples after six months of storage at -20°C and -70°C.

Conclusion: This NGS-based platform offers a sensitive, accurate, and efficient solution for detecting lentiviral integration sites across various sample types, with a detection limit as low as 0.5 ng/µL. It maintains stability after long-term storage and incorporates UMIs, error correction, and dual size selection to ensure high specificity and reproducibility. Validated by Sanger sequencing and supported by an automated, cloud-based bioinformatics pipeline, the platform is well-suited for both preclinical and clinical gene therapy applications, enabling reliable integration site analysis at scale.

References: Chakravarty, Debyani, et al. "OncoKB: a precision oncology knowledge base." JCO precision oncology 1 (2017): 1-16.
Hamada, Motoharu, et al. "Integration mapping of piggyBac-mediated CD19 chimeric antigen receptor T cells analyzed by novel tagmentation-assisted PCR." EBioMedicine 34 (2018): 18-26.
Rosewick, Nicolas, et al. "An improved sequencing-based bioinformatics pipeline to track the distribution and clonal architecture of proviral integration sites." Frontiers in microbiology 11 (2020): 587306.
Sheehan, M., Kumpf, S. W., Qian, J., Rubitski, D. M., Oziolor, E., & Lanz, T. A. (2024). Comparison and cross-validation of long-read and short-read target-enrichment sequencing methods to assess AAV vector integration into host genome. Molecular Therapy Methods & Clinical Development, 32(4).

Acknowledgements: This study was funded by Accurant Biotech Inc. The authors declare no competing interests. All procedures involving human or animal materials were conducted in accordance with relevant ethical guidelines and approved protocols.


Experimental Design for Library Preparation. DNA was sheared by sonication, followed by adapter ligation and two rounds of PCR to add sequencing linkers. First-round PCR use primers specific to lentiviral long terminal repeat (LTR) and adapter sequences to amplify integration sites.


Chromosomal Distribution of Integration Sites. Each red arrow or line indicates the specific genomic location of a detected integration site across the chromosomes.


Sensitivity result. A series of DNA library dilutions ranging from 0.05% to 5% were prepared. The assay demonstrated a sensitivity of 0.05%, enabling the detection of low-frequency insertion events. Linear regression analysis showed an R² value greater than 0.997, indicating strong assay performance and accuracy.