Tumor-associated macrophages (TAMs) are key mediators of tumor immune evasion. However, their regulatory circuits and checkpoints are partially understood. Here, we generated a TAM regulatory network by integrating human tumors single-cell RNA sequencing (scRNA-seq) data with a dedicated CRISPR screen. Using a deep generative model, we constructed a gene perturbation network linking individual candidates with prototypical TAM functions. We identified Zeb2 as the master regulator of TAM programs, orchestrating suppression of type-I interferon response and antigen presentation alongside activation of immune suppression programs. Genetic ablation of ZEB2 reprograms TAM function and identity on the chromatin, RNA, and protein levels. In macrophage-rich human tumors, ZEB2 expression is associated with poor prognosis. Selective Zeb2 in vivo targeting reprograms TAMs and mobilizes systemic T cell responses, achieving robust tumor clearance. Overall, our study generates a detailed roadmap of TAM gene circuits and identifies ZEB2 as a master switch with therapeutic potential.
Keywords: CRISPR; TAM reprogramming; Zeb2; cancer immunology; cancer immunotherapy; deep generative modeling; regulatory network; single cell genomics; systems immunolgy; tumor-associated macrophage.
Copyright © 2025 Elsevier Inc. All rights reserved.