Abstract: The ability to selectively control transcript splicing in response to endogenous RNA signals represents a major frontier in synthetic biology. Synthetic RNA devices that detect specific transcripts can, in principle, distinguish target cells based on their molecular signature and execute specific functional outputs. As part of the SmaRT (Self-activating mRNA transcripts) project, we present a quantitative framework and proof-of-concept study using engineered Tetrahymena Group I ribozymes to achieve programmable, RNA-dependent splicing activation in mammalian cells.
To optimise the system, we first identified the ideal cis-splicing configuration through a rigorous in silico screening pipeline. Candidate intron insertion sites within the NanoLuc coding sequence were evaluated using base-pairing probability matrices to prioritise minimal exon-intron invasion, structural preservation of the intron core, and optimal splice-site proximity within the folded RNA. This computational selection was followed by experimental validation in HEK293T cells. Building on this, we designed and screened various split-sites within the ribozyme sequence to determine the most efficient trans-splicing configuration, serving as our base construct for our programmable sensors. Next, we created a conditionally activated ribozyme architecture operating with only two RNA species: a single sensor RNA and a cognate activator RNA. In this bipartite system, the engineered intron is maintained in an unactivated state by default. Sensing the target transcript induces a conformational shift that triggers the ribozyme, restoring a functional open reading frame for downstream translation.
This work establishes a foundational proof-of-concept and quantitative framework for cell-type-specific mRNA splicing control, laying the groundwork for future applications in diagnostic RNA sensing, conditional mRNA therapeutics, and programmable synthetic-biology circuits.