BME PhD Dissertation Defense: Jaehoon Choi

Title: "Signaling motif refactoring uncovers positional and receptor expression dependency of ITIM in chimeric inhibitory receptor"

Advisory Committee: Wilson Wong, PhD – BME (Research Advisor) Mark Grinstaff, PhD – BME (Chair) Juan Fuxmann Bass, PhD – Biology Brian Cleary, PhD – CDS, BME, Biology Ahmad (Mo) Khalil, PhD – BME

Abstract: Solid-tumor CAR T-cell therapy remains constrained by on-target off-tumor toxicity because most available antigens are tumor-associated rather than truly tumor-specific. NOT-gated strategies using inhibitory chimeric antigen receptors (iCARs) offer a promising route to improve tumor-versus-normal discrimination, but the field still lacks both a quantitative framework for measuring inhibitory function and mechanistic design rules for engineering inhibitory domains. In this dissertation, I established quantitative platforms to measure inhibitory functions of iCARs in Jurkat (Chapter 1), examined how leukocyte immunoglobulin-like receptor 1 (LIR1)-based iCAR functions mechanistically (Chapter 2), and finally discovered the grammar of immunoreceptor tyrosine-based inhibitory motifs (ITIMs), classic inhibitory motifs, of LIR1 iCAR (Chapter 3), enabling motif-level iCAR engineering. Combinations of an identified high-dynamic activating CAR Jurkat reporter clone and reliable workflow of lentiviral production and Jurkat transduction in 96-well plates provide the experimental foundation for the following mechanistic and design-rule studies. Using phosphatase-deficient Jurkat models and bystander assays, I showed that LIR1-mediated inhibition is strongly SHP-1 dependent and functions primarily in cis under the tested conditions, while also showing that neither non-ITIM/ITSM (switchable motif) inhibitory tails nor dual-domain stacking strategies generalize cleanly as iCAR engineering solutions. I built a rational 43-construct LIR1-based iCAR library to interrogate motif identity, copy number, and position using the high-dynamic Jurkat assay. This approach revealed a clear motif hierarchy, non-linear saturation with increasing intact ITIM number, motif-dependent pairwise interaction, an ITIM2-ITIM4 paradox that distinguishes amplifier-like from stabilizer-like inhibition, and motif-specific positional dependency. Primary human T cells preserved the dominant parts of this grammar. These findings help convert inhibitory CARs from empirically assembled receptors into systems governed by a definable inhibitory motif grammar.