Identification and validation of novel human genomic safe harbor sites

Abstract

Numerous gene addition methods are gaining increasing popularity in the field of gene therapy, where replacement of the mutated copy of the gene is required, as well as in cell engineering, in which synthetic receptors can be introduced into a cell or a group of cells to create artificial gene circuits capable of eliciting therapeutic or tissue enhancing functions. Existing gene addition tools suffer from heterogeneity of transgene expression levels and may cause aberration to normal transcriptomic profile due to up- or down-regulation of both protein coding and non-protein coding genes. With the advent of targeted gene integration methods, the necessity for the identification of genomic loci, which would support durable and safe transgene expression – Genomic Safe Harbor (GSH) sites – became ever more urgent. In this dissertation I describe a pipeline for computational prediction and experimental validation of novel human GSH sites using existing as well as newly introduced genomic safety criteria. In chapter 1 I explain the use of a rational approach to verify computationally predicted genomic sites by targeted integration of reporter as well as therapeutic genes into select computationally predicted locations. This approach yielded the identification of two candidate GSHs, which showed robust and durable expression in investigated cell lines and were later confirmed in primary human T cells and primary human dermal fibroblasts. The safety of transgene expression upon integration into these two sites was subsequently verified using bulk and single-cell transcriptomic analyses, which showed minimal changes in global RNA expression levels following transgene integration. Overall, these two newly identified GSH sites create a broad platform for safer and more reliable gene addition-based gene and cell therapies, facilitating their transition into clinical practice. In chapter 2 I describe an attempt to implement a multiplexed experimental search of novel GSHs using high-throughput library-based approach. Specifically, described method would allow for a rapid screen of thousands GSH sites exploiting a library of guide RNAs targeting various computationally predicted GSH locations and a non-homologous end joining pathway to drive targeted insertion of a reporter transgene into a genomic locus determined by a guide RNA library member. Such pooled approach would allow to reveal a set of highly transcribed loci, allowing for their subsequent validation by individual transgene integration and transcriptomics assessment. This study, however, was associated with numerous experimental hurdles and was eventually discontinued with suggestions on further optimizations in the future. To date, only three empirically validated sites in the human genome have been reported for durable expression in different cellular contexts. However, all three of them are located in gene dense regions surrounded by proven oncogenes, significantly increasing the risk of integration-induced tumorigenesis. Furthermore, they do not support the rapid pace of innovation in synthetic biology that enables multiple transgene integration and genetic circuits to rewire and reprogram cellular function. Two novel, computationally and experimentally validated GSH sites described in this thesis open new opportunities for safer and more predictable genome engineering of human cells, expanding the toolkit for diverse cell therapy and synthetic biology applications, from the treatment of inherited disorders by replacing mutated genes with their functional copies, to creating synthetic networks in immune cells to drive multi-input response, to augmenting properties of cells and tissues by safe addition of enhancing transgenes. Finally, thanks to long-term high levels of transgene expression, identified GSH sites can be used for large-scale therapeutic protein manufacturing in human hosts.