1119 Pool discovery of chimeric antigen receptors to program macrophages

<h3>Background</h3> Chimeric antigen receptors (CARs) have been primarily designed based on endogenous T-cell molecules, such as CD3 and CD8, for over 35 years. However, the suitability of T-cell-based CAR structures for other immune cell types remains unclear. Notably, CD28 and 4-1BB CARs, which perform well in T-cells, exhibit compromised protein trafficking in macrophages. This prompted us to explore optimal CAR structures for macrophages. <h3>Methods</h3> We conducted a pooled library screening of CAR structures incorporating macrophage-specific endogenous signal peptides, hinge regions, transmembrane, and signaling domains.¹ Using a modular cloning strategy, we generated a library of 50,000 CAR variants for functional screening. Our FlowSeq approach enables the identification of CARs that enhance phagocytosis and cytotoxic cytokine secretion. <h3>Results and Conclusions</h3> To further refine our findings, we employed neural networks to infer a larger library and define the most effective CAR structures for macrophages. This high-throughput strategy, combined with advanced neural network analysis, paves the way for the discovery of novel CARs tailored for immunotherapy. <h3>Acknowledgements</h3> This research is supported by Health@InnoHK, the Innovation and Technology Commission of the Government of HKSAR. <h3>Reference</h3> Ronan W O’Connell, Caleb J Bashor. Ultra-high throughput mapping of genetic design space. <i>bioRxiv</i>[Preprint]. 2023 Jul 6:2023.03.16.532704.