Abstract 4072: Functional screen to optimize logic gate potency and selectivity

Abstract Background: Logic-gated cell therapy systems are considerably more complicated than conventional therapeutics and, therefore, more challenging to optimize. One type of logic gate, the NOT gate, responds to two inputs: the presence of antigen A and absence of antigen B. Perhaps the best studied NOT gate is called Tmod which consists of an activator (eg, a chimeric antigen receptor [CAR]) that is co-expressed with an inhibitory receptor, or blocker (DiAndreth B, et al. Clin Immunol. 2022). However, the complexity of the system limits the amount of structure-activity information that can readily be obtained. If further modules (eg, potency boosters) are added to the system, the behavior of these synthetic circuits can be even more complex. Methods: A fluorescence-activated cell sorter (FACS)-based screening system was designed to rapidly select functional inhibitory receptors and inducible boosters from a pooled library of candidates. This screen incorporates an NFAT-responsive reporter cell line for iterative selection of reporter (+) and (-) cell populations by FACS after exposure to different stimuli. The selected populations were analyzed by sequencing to identify candidate blocker or booster sequences enriched during the process. Results: Proof-of-concept experiments using a library of blocker variants demonstrated enrichment of active blockers. Moreover, selected features were used to successfully predict the blocker behavior with 70-80% accuracy by machine-learning algorithm. In another set of experiments, a library of almost 30,000 variants of a signal 1 booster that activate signaling in the presence of a small molecule (rimiducid) was screened. Signal 1 boosters with specific domain structures were enriched and validated in subsequent experiments. Conclusions: The screening system described here identifies active blocker receptors and boosters from large libraries of variants. Its throughput is sufficient to generate large datasets and train machine learning models to better predict and optimize the function and control of logic-gated cell therapeutics. Citation Format: Sara Martire, Michele Mcelvain, Vasantika Suryawanshi, Mark Daris, Tanveer K. Gill, Timothy P. Riley, Alexander Kamb, Chawita Netirojjanakul. Functional screen to optimize logic gate potency and selectivity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4072.