Abstract Cell and gene therapies against cancer could be empowered by sensors that produce proteins in response to specific intracellular markers (e.g., expressing a cytotoxic payload in response to oncogenes) or extracellular cues (e.g., expressing a chimeric antigen receptor only when T cells are in the tumor microenvironment). An emerging delivery modality is in vitro transcribed mRNA, which holds the promise of superior safety and affordability than more conventional DNA vectors. While synthetic biology tools have predominantly relied on DNA-level regulations, here I will share two sensor designs uniquely compatible with mRNA delivery. First, for intracellular markers, a promising strategy is sensing mRNA transcripts. It is now straightforward to obtain single-cell transcriptomics, which are highly informative of cell type/state. If we can build RNA sensors to output arbitrary proteins in direct response to specific RNAs, we will access virtually any cell type/state for research and therapy. Furthermore, RNA base pairing would facilitate programmable designs. Previous RNA sensors relied on base pairing-induced conformational changes of RNA, and have met with limited success in mammalian cells. As an alternative strategy, we took advantage of endogenous ADAR (adenosine deaminases acting on RNA) enzymes that specifically edit adenosines to inosines (treated as guanosines during translation) in dsRNA, and created RADAR (RNA sensing using ADAR, Kaseniit et al., 2023). RADAR relies on the formation of dsRNA between the sensor RNA we introduce and the endogenous input RNA we’d like to detect; ADAR then edits a strategically positioned stop codon (UAG) into a tryptophan codon (UIG), enabling the translation of the downstream output coding sequence. We validated and optimized RADAR, and reported its compatibility with human/mouse transcriptomes. We showed that RADAR is programmable and modular. It also uniquely enables compact implementation of AND logic, important for integrating the inputs from multiple markers and unambiguously identifying cells of interest. We have since demonstrated RADAR in a variety of contexts including cancer lines with an elevated oncogene. Second, while transcripts mark cell types/states, extracellular ligands reflect the other key aspect that payload production could be conditioned on – the microenvironment. Although our field has plenty of synthetic receptors to offer when it comes to sensing ligands, all previously engineered modular ones require DNA-level operation. Inspired by the power of ADAR editing, we created LIDAR (Ligand-Induced Dimerization Activating RNA editing, Zhang & Mille et al., 2024). LIDAR brings an attenuated ADAR to the proximity of a stop codon in a ligand-dependent fashion, and is compatible with diverse inputs, outputs, and cellular contexts. We are in the process of demonstrating RADAR and LIDAR to improve cell therapies and enable non-invasive diagnostics. Citation Format: Xiaojing Gao. Programmable RNA sensors for internal states and external cues in living cells [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: RNAs as Drivers, Targets, and Therapeutics in Cancer; 2024 Nov 14-17; Bellevue, Washington. Philadelphia (PA): AACR; Mol Cancer Ther 2024;23(11_Suppl):Abstract nr I006.
Read full abstract