Genetic mutations in HSV-1 replication-defective vectors: Implications for their safety in gene therapy applications

Beyond its well-known role in orofacial recurrent infections, HSV-1 has garnered significant attention in neuroscience for contrasting reasons. On one hand, it has been found to be involved in neurodegenerative processes; on the other, it may represent a versatile platform for gene therapy of brain diseases, due to its large genome that enables the delivery of sizable or multiple genes. These opposite features underscore the importance of understanding HSV-1 interactions with neural tissues in view of its employment as a gene therapy platform. We recently developed a new generation of highly defective backbones that proved very efficient and safe after direct injection in the brain parenchyma. Here we aimed at probing in depth the safety of viral batches that lack obvious unwanted (specifically, fusogenic) activities during production and, therefore, may escape negative selection. We employed whole-genome sequencing, electrophysiology, and viral engineering to compare different viral batches. We identified mutations (in particular A to I at position 549 in the UL27 gene) that confer fusogenic capacity to the envelop glycoprotein gB, inducing a hyperexcitable phenotype in transduced neurons. Such syncytial variants should be identified and avoided for any application of HSV-1 vectors implicating their direct injection in the nervous system.