P09 Next-generation sequencing: on-and off-target analysis of gene editing in fibroblasts from patients with recessive dystrophic epidermolysis bullosa

Abstract Introduction and aims Base editors (BEs) and prime editors (PEs) have proved to have advantages over traditional clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9 in correcting pathogenic mutations. We previously demonstrated 94.6% efficiency using adenine base editor (ABE)8e in correcting nonsense mutation (c.5047 C>T) in fibroblasts, which causes recessive dystrophic epidermolysis bullosa (RDEB). However, safety profile and cell-type-specific efficiencies of gene editing techniques have not been characterized. The aim of our research is to establish corrections of c.5047 C>T pathogenic variant in fibroblasts and compare off-target effects and efficiency of CRISPR-Cas9, BE and PE systems. Methods Primary fibroblasts harbouring heterozygous C>T pathogenic variant on exon 54 will be transfected with CRISPR-Cas9, ABE8e and PE using electroporation or lipofectamine. Both Sanger sequencing and next-generation sequencing (NGS) will be conducted to confirm pathogenic variant corrections. Galaxy platform which incorporates different NGS analysis tools will be used to analyse the NGS data. Subsequently, whole-exome sequencing and targeted amplicon sequencing will be used for off-target analysis. Results On-target efficiency for ABE8e to correct c.5047 C>T pathogenic variant in patient primary fibroblasts was determined first by Sanger sequencing showing 100% editing in contrast to 84% determined by NGS. Moving forward, we plan to correct the same pathogenic variant with PEs. Initially experiments are under way with PEmax mRNA and engineered PE guide RNAs (epegRNA) to compare different PEs. Furthermore, we will investigate the off-target effects of CRISPR-Cas9, ABE8e and PEs to edit RDEB primary fibroblasts using NGS Methods and conduct functional studies. Conclusions Here, we showed that the c.5047 C>T pathogenic variant in RDEB fibroblasts could be successfully edited using ABE8e. We demonstrated that NGS is required to precisely determine the editing efficacy. Taken together, we have established a pipeline for base editing in RDEB and we aim to achieve the same for prime editing in the future.