Abstract:
RNA editing has emerged as a powerful strategy for manipulating gene function at the transcript level. In contrast to permanent genome editing, site-directed RNA editing enables precise and reversible changes in RNA. This transient modulation genes and thereby of protein function without altering the DNA, makes site-directed RNA editing particularly attractive both as a research tool and as a therapeutic modality. Among the available RNA editing platforms, SNAP-ADAR provides a versatile system in which chemically modified guide-RNAs recruit SNAP-tagged ADAR deaminases to a defined transcript, enabling selective A-to-I editing. Since guide-RNAs can be designed for different target transcripts based on their sequence, SNAP-ADAR system is well suited for systemic interrogation of endogenous gene function. This is particularly relevant in the liver, a central metabolic and immunological organ whose signaling influence not only hepatic inflammation, regeneration, nutrient handling, but also homeostasis across the organism. Hepatic pathways are dynamic and many disease-relevant phenotypes arise not from complete loss of a gene function, but from altered activity of specific regulatory networks, metabolic or inflammatory programs. In this setting, RNA editing offers an attractive means of perturbation, as it can be used to introduce defined changes in coding regions and prove functional consequences in a controlled and reversible manner. A robust liver-compatible SNAP-ADAR platform could serve as a tool for mechanistic studies, but also as a basis for target screening and prioritization.
This thesis aims at advancing the SNAP-ADAR platform by optimizing guide-RNA designs in order to bridge the gap between efficient in vitro editing and practical, liver-compatible delivery format suitable for future in vivo screening applications. On the foundation of SNAP-ADAR-version-1, version-2 was developed to boost editing by extending the gRNA length (22 to 25 nt) and introducing LNA modification to strengthen target hybridization. Further, it transitioned from a monovalent benzylguanine (BG) linker strategy to bivalent Bis-BG linker, enabling recruitment of two SNAP-ADAR molecules per guide-RNA. Version-2 achieved strong editing across a broad panel of approximately 50 targets in cell culture, but was constrained by poor serum stability and limited manufacturability due to branched linker synthesis and non-scalable coupling. These limitations were incompatible with triantennary N-acetyl galactosamine (GalNAc3)-based liver delivery, where substantially higher guide-RNA material is required. This thesis contributed partly to version-2, but focused primarily on the development of version-3.
To overcome these limitations, the platform was further optimized to generate version-3, a format intended to preserve high editing performance while improving stability and scalability. In version-3, the branched bivalent linker was replaced with two simple monovalent linkers integrated into a redesigned guide-RNA, preserving functional equivalence while enabling straightforward upscaling. In parallel, a systematic chemical optimization identified a highly sensitive region adjacent to the central base triplet, in which specific phosphorothioate (PS) incorporation markedly improved nuclease resistance without sacrificing the editing yields. The resultant, version-3 design was compact, stable and non-toxic in cell culture. In delivery-focused experiments, PS-optimized version-3 guide-RNAs performed better following GalNAc3-mediated uptake in primary mouse hepatocytes compared to the unoptimized intermediate version-3 or the long version-2 guide-RNAs. Further optimization produced gRNA-657 (lacking 5’ wing LNA), achieving ~34% editing of mStat1_Y701C at 1 µM (vs ~28% for the PS-optimized predecessor) with minimal guide-RNA-dependent off-target editing as analyzed by next-generation sequencing. Because STAT1 is a central regulator of inflammatory signaling, these findings provide an initial proof-of-concept for ex vivo hepatic editing. Overall, this work delivers a streamlined, scalable and stability-optimized SNAP-ADAR guide-RNA platform that closes key gaps between efficient in vitro editing and liver compatible delivery designs, establishing a practical foundation for future in vivo target identification.
RNA editing