Decoding Immune Cell Communication with de novo Designed Cytokines

Abstract

Natural cytokines are powerful immune signaling molecules, but their pleiotropic nature limits their therapeutic potential because they trigger diverse, and often conflicting, responses across different cell types. During my graduate research, I developed computational and experimental approaches to redesign cytokines from the ground up, creating synthetic signaling molecules with programmable, cell-type-specific activity.This work established two independent design strategies for controlling cytokine function. First, I developed a computational design pipeline that scaffolds receptor-binding motifs into defined spatial geometries, demonstrating that receptor geometry alone can tune signaling outputs without altering binding affinity. Second, a large-scale screen of more than 1,000 non-natural receptor combinations uncovered a fundamental rule governing receptor compatibility: productive signaling requires pairing a broadly compatible “common” receptor with a selective “private” receptor that determines response specificity. This principle enabled the identification of tens of previously unknown signaling pairs, which were tested in immune and cancer cell models, demonstrating the broad versatility of this design framework. Together, these studies establish independent control over receptor affinity, geometry, and receptor pairing, providing a general framework for engineering synthetic cytokines with precise cellular specificity. This work lays the foundation for programmable immunotherapies that selectively target desired cell types while minimizing off-target effects.