Decoding the Determinants of T Cell States: From Systematic Genetic Screens to Scalable Tissue Models

Abstract

A central challenge in immunology is understanding how individual cell programs translate into coordinated tissue-level responses. Dissecting these processes requires tools that can measure and perturb both internal genetic logic and external cellular context. Current methods are often lacking in their ability to scale across the thousands of potential genes and regulatory elements that govern immune cell phenotype. This work addresses these limitations through two complementary projects: first, transcription factor wide screens identifying novel regulators of T cell fate and function, and second, a scalable in vitro 3D tissue culture platform to profile local cell-cell interactions.In the first project, we performed transcription factor-wide overexpression (OE) screens to identify regulators of T cell memory and tissue residence. This uncovered an unrecognized role for the TGIF family in programming T cell state. Specifically, we found that TGIF2LX overexpression increases the expression of key markers including IL7R and CD103. While shRNA-mediated knockdown of this pathway revealed its necessity for robust anti-tumor immunity, we found that its expression must be tightly regulated, as constitutive overexpression ablates tumor control. Mechanistically, we show that the TGIF family regulates a ciliary gene program that T cells repurpose to control cell surface marker abundance. We show TGIF2LX OE synergizes with tissue-residence promoting factors (TGF-b and retinoic acid) to upregulate markers associated with homing, adhesion, and retention. Next, we address the challenge of probing how complex cellular networks change in response to perturbations. While single-cell RNA-sequencing (scRNA-seq) offers deep transcriptomic insights, the required tissue dissociation typically results in the loss of local spatial context. To overcome this, we leverage a microfluidic technology that captures heterogeneous cellular suspensions within thousands of discrete 3D hydrogel droplets. In this work, we present IMPACT (Indexing Microtissue Parts with Antibody-based Combinatorial Tagging), a method that utilizes combinatorial indexing to label cell populations within each encapsulated microtissue with a unique set of sequencing-compatible barcodes. This enables the computational reconstruction of individual microtissues from single-cell data, preserving the link between a cell’s state and its local neighborhood. As a proof-of-concept, we applied IMPACT to malignant ascites-derived microtissues, uncovering how local immune cell interactions are reshaped in response to treatment. Together, these works establish a scalable framework for systems immunology, bridging the gap between discovery of intracellular genetic drivers and characterization of the multicellular environments in which they operate.