Searching for new barrier insulator parts for mammalian synthetic biology

Abstract

In any biotechnological application that requires a transgene to be expressed long-term in mammalian cells, preventing its silencing by epigenetic processes can be a challenge. This includes the biopharmaceutical production of therapeutic proteins in Chinese Hamster Ovary cells and gene therapies currently being trialled for diseases such as Severe Combined Immunodeficiency. One of the strategies available for ensuring the stable expression of a trans-gene from the mammalian genome is to include chromatin control elements, such as barrier insulators, in the vector it integrates with. These sequences act in cis to maintain nearby genes within an epigenetic environment permissive to transcription. Barrier insulators are not a particularly well-catalogued variety of genetic part, so new examples could help improve our understanding of how they function and are desired for building genetic circuits. This thesis discusses an effort to develop a high-throughput barrier insulator testing platform, based on a Bxb1 integrase landing pad. Landing pads are loci in a genome that allow highly efficient targeted integration of DNA cargoes by site-specific recombination. In this case, the landing pad would be used to introduce many prospective barrier insulators into the genome of a human cell line, where they would then be evaluated. Barrier activity would be assessed at the landing pad using an epigenetic silencing assay that would follow the expression of a fluorescent reporter gene. The reporter would be silenced experimentally by tethering repressor domains, involved in heterochromatin formation, to a site downstream in the genome. This should seed the formation of heterochromatin that should spread into, and repress the transcription of a reporter. Placing a functional barrier insulator in the path of this spreading has previously been shown to mitigate silencing. Chapter 3 describes the development of an experimental setup for inducing and monitoring gene silencing in the genome of HEK293-FT cells. In this work, a cell line containing a CAG driven EGFP reporter and downstream TetO7 recruitment site at the AAVS1 locus was generated. It was found that targeting several Streptococcus pyogenes (Sp) dCas9 repressors to the TetO7 site could effectively silence EGFP expression within six days. This chapter also describes the development of Ouabain co-selection and Double Cutting (ODC), a CRISPR-Cas9 method for targeted integration, which proved to be an effective means for generating edited HEK293-FT cell lines. In chapter 4, the silencing assay was utilised to screen for new CRISPR repressor tools based on the CRISPR systems from Staphylococcus aureus (Sa dCas9), the commercially relevant Eubacterium rectales (Er dCas12a), Acidaminococcus sp. BV3L6 (As dCas12a) and Lachnospiraceae bacterium ND2006 (Lb dCas12a). It was anticipated that these might be applied as complementary, orthogonal repressors to Sp dCas9 repressors. Three novel repressors were identified that are able to silence EGFP expression: As dCas12a-csHP1a, Sa dCas9-KRAB-MeCP2 and Sa dCas9-REPR. Lb dCas12a and Er dCas12a repressors appeared to be too poorly expressed to detect any silencing effects. Er dCas12a, despite not being suitable for repression, was demonstrated to be suitable as a scaffold for activating gene expression for the first time. In chapter 5, the experimental set up from chapter 3 was used to inform the design of a Bxb1 landing pad platform for testing insulators. This design utilised a promoter trap for selecting cells which had undergone Recombinase Mediated Precision Integration (RMPI). HEK293-FT cell lines containing a single copy of the landing pad were isolated, and a set of cargo vectors that could recombine into the landing pad were generated. These vectors each contained a promoter-less mNeonGreen reporter cassette, a different insulator sequence and the TetO7 array. Unexpectedly, when the cargo vectors were recombined into landing pad cell lines, the recombinant cells were found to express different amounts mNeonGreen depending on the insulator sequence they now harboured. This hampered the planned epigenetic silencing assay, and the cause of this phenomenon could not be conclusively determined. Despite this complication, two of the novel candidate insulator sequences introduced into the landing pad exhibited characteristics that could indicate barrier activity. Both ensured high levels of mNeonGreen expression from the landing pad up to 42 days post-transfection. This was comparable to the level observed from a cargo plasmid containing two copies of the well-characterised cHS4 barrier insulator sequence. In summary, this work made progress towards developing a high-throughput insulator testing platform. Two sequences linked to long-term enhanced trans-gene expression were identified, which will require further exploration and validation. Along the way, new CRISPR tools for controlling transcription based on the As dCas12a, Er dCas12a and Sa dCas9 scaffolds were also demonstrated. Further pursuing this barrier assay may yield additional sequences with insulator activity. This may eventually help to determine the design rules necessary for building improved, synthetic barrier insulator sequences de novo.