Genome sequencing for evaluating the efficacy, specificity, and safety of human genome editing

The goal of this proposal is to develop a comprehensive approach for evaluating the efficiency and specificity of genome-edited human cells using whole-genome sequencing. Genome editing has enormous therapeutic potential by making it possible to restore genetic defects, inactivate deleterious mutant alleles, and augment the function of cellular therapies. Although genome editing technologies are designed for optimal efficiency and specificity, on-target editing can be variable, and unwanted mutations in edited cells can result in unintended functional consequences, including disruption of genes due to off-target mutations, transgene insertions, or deletions, duplications, or structural rearrangements. As a result, current draft guidance from the Food and Drug Administration (FDA) recommends that genome-edited cellular therapies be evaluated for both on- and off-target mutations. However, existing approaches for performing these analyses are logistically complicated and either use antiquated methods or involve modifications to the editing process that cannot be applied to cellular drug products that will be used in patients. We hypothesize that whole-genome sequencing (WGS) is an ideal platform to address FDA guidelines for genomic analysis of genome-edited cellular products because it detects the full spectrum of mutation types and can be used to evaluate fully manufactured ‘patient-ready’ cellular therapies. Here we propose to develop a comprehensive WGS assay specifically designed to characterize mutations in genome-edited human cells. In Aim 1, we will modify our recently developed clinical WGS assay for somatic mutations (ChromoSeq) to measure the efficiency and specificity of genome editing in human cells. We will use high coverage (>250x) WGS of paired edited and unedited control cells and joint somatic variant calling methods to quantify on-target editing efficiency and detect transgene integration sites and unintended single nucleotide variants, insertions/deletions (indels), and chromosomal rearrangements. We will then qualify this WGS approach using a dataset of high confidence mutations generated in three human cell lines with CRISPR/Cas9 and multiplex pools of guide RNAs (gRNA), which will be identified via iGUIDE and targeted, error-corrected deep sequencing. In Aim 2, we will use our WGS assay to define the landscape of mutations in genome-edited human CAR-T cells. These will include 5 replicate experiments with reagents to common CAR-T targets, and 15 existing primary human CAR-T products edited at a range of therapeutically relevant genes that have already been generated in our labs. We will use these data to generate a benchmark dataset of on-target editing efficiency measurements, CAR integration sites, and unintended mutations in human CAR-T cells that will provide valuable data for future clinical trials. Finally, we will analyze up to 20 additional genome-edited cellular products from the Somatic Cell Genome Editing Consortium to further establish the performance and utility of WGS for evaluating the safety and efficacy of genome-edited cellular therapies that will enable future investigational clinical studies.