Optimization of a large template-driven gene editing strategy for Epidermolysis Bullosa Simplex: SV40 versus BGH polyadenylation signals for functional keratin 14 expression.

Epidermolysis bullosa simplex (EBS) is a rare inherited skin disorder characterized by skin fragility and blistering. It is primarily caused by dominant mutations in the KRT5 and KRT14 genes, encoding keratin 5 (K5) and keratin 14 (K14), respectively, which together form the intermediate filament network of basal keratinocytes. Currently, therapeutic options for EBS remain largely palliative, underscoring the urgent need for durable treatments. This study aims to develop a mutation-independent, universal CRISPR-Cas9 gene editing strategy capable of correcting the entire spectrum of KRT14 mutations in primary human keratinocytes via homology-directed repair (HDR). To achieve this, we employed a large adeno-associated virus (AAV)-delivered template encoding a healthy KRT14 sequence. To optimize transgene expression, we evaluated two distinct polyadenylation (pA) signals, Simian Virus 40 (SV40) and Bovine Growth Hormone (BGH), for their ability to drive functional expression of the therapeutic keratin 14 (K14). Primary keratinocytes derived from an EBS patient were edited using sgRNA-Cas9 ribonucleoprotein complex targeting intron 1 of the KRT14 locus, combined with AAV6 donor templates carrying a promotorless, codon-optimized, full-length KRT14 coding sequence. To maximize editing efficiency and increase the likelihood of targeting the epidermal stem cell compartment, small molecule inhibitors of the competing non-homologous end joining (NHEJ) and microhomology-mediated end joining (MMEJ) pathways were utilized. Under standard conditions, high editing efficiencies of up to 50% were achieved, which increased to 80% under pharmacological enhancement of HDR. Moreover, dual pathway inhibition increased the frequency of biallelic targeting, promoting structural restoration of a wild-type-like intermediate filament network. Although, molecular characterization confirmed precise on-target integration, it also revealed the presence of concatemeric and NHEJ-mediated aberrant integration events at the target locus, which persisted despite pharmacological inhibition. Importantly, these genomic alterations did not result in detectable aberrant protein byproducts. Protein expression analysis demonstrated the superior performance of the BGH-pA signal over SV40-pA. Whereas SV40-pA resulted in lower expression levels and putative transcriptional readthrough, BGH-pA ensured robust, physiological expression of the therapeutic K14 and efficient transcriptional termination. In conclusion, this study demonstrates the highest reported efficiency for large template-driven gene editing in primary human keratinocytes and supports the development of a universal gene editing approach for KRT14-associated EBS, laying the groundwork for a scalable therapeutic strategy that could benefit multiple patients. Furthermore, these findings highlight the critical role of 3' regulatory element optimization to achieve therapeutic transgene expression, identifying BGH-pA as a key element for promoting functional K14 expression, thereby advancing the development of a definitive gene therapy for EBS patients. Ongoing and future studies will prioritize comprehensive genotoxicity profiling and the validation of phenotypic rescue in advanced 3D skin models to ensure long-term safety and efficacy prior to clinical translation.